Closed-Loop Control by Laser Power Modulation in Direct Energy Deposition Additive Manufacturing

Abstract

Direct Energy Deposition is a metal additive manufacturing technique that has raised great interest in industry thanks to its potential to realize complex parts or repairing damaged ones, but the complexity of this process still requires much effort from practitioners to achieve functionally sound parts. One of the recurring flaws of such parts is the phenomenon of over-deposition, which may occur due to unpredicted local increases of energy density. The deposition of uniform metal tracks is critical in many practical cases, when parts are composed by a significant number of layers and/or when complex tool paths induce heat build-up, for example in thin structures. Therefore, detecting anomalies such as over-growth in real-time and dynamically correcting them is of paramount importance for achieving repeatable, first-time-right parts. This work studies the use of a closed-loop control system for Direct Energy Deposition, proposing to adjust on-line the power of the laser beam according to the feedback provided by the analysis of melt pool images. The images are acquired by a camera, mounted coaxially into the optical chain of the deposition head, which records images at 100 fps while the process is running. The proposed approach is explored experimentally by comparing the over-deposition measured on sample test geometries obtained with a traditional feed-forward approach with the over-deposition obtained through the developed closed-loop control laser deposition system.

Controlling grain structure in metallic additive manufacturing using a simple, inexpensive process control system

“…Melt pool area requires a contour to be plotted at a threshold value which is computationally expensive; maximum intensity takes a single value, so is prone to error if any sensor anomalies occur. For these reasons, the thermal intensity was used as the measure of "thermal conditions" to be controlled, similarly to Baraldo et al 21 .…”

“…The resultant 12bit images (500 x 500 px, where 1 px = 5.4 µm) were analysed in MATLAB R2021b (Mathworks Inc). For each image, all the pixels were summed to give an overall thermal intensity as previously reported 21,35 ; this is a sum of pixel digital levels, so is unitless. This was compared to both the maximum intensity in a single image and to the melt pool area, both showing strong correlation below a thermal intensity of 0.2x10 7 34 , below which, the laser power was ramping at the ends of each hatch.…”

Multi-Faceted Monitoring of Powder Flow Rate Variability in Directed Energy Deposition

“…The coaxial camera was used to capture images of the melt pool. The greyscale values of all pixels in each image were summed, to give a single 'brightness' value for each image [22] . This is not a measurement of melt pool size, but it is expected to show the same general trends as melt pool size within a single build.…”

Understanding the effects of temporal waveform modulation of laser emission power in laser powder bed fusion: Part II - Experimental investigation

“…Xiao et al [18,19] have shown that, for powder-blown directed energy deposition of IN718, modulation of the power emission in the millisecond range can have a positive effect in improving both elongation and yield strength due to the reduced formation of the Laves phase. Recent work by Baraldo et al [20] also showed how conventional laser-power modulation may be employed to build a control architecture.…”