Improving SLM additive manufacturing operation precision with H-infinity controller structure

https://doi.org/10.1016/j.cirpj.2020.09.007Get rights and content

Abstract

SLM (Selective Laser Melting) is the most widespread additive manufacturing technique of metal part. The desired part is elaborated through local melting of a raw metal powder bed by means of laser. In industrial machines, galvanometer motors achieve the laser beam deflection and focus control tasks. This paper proposes a H-infinity controller synthesis which improves the system accuracy and robustness towards physical features. Compared to a conventional control scheme, results obtained with the H-infinity controller implemented in an open architecture test bench consisting of a 2-axis laser deflection system showed improved accuracy performance while operation rapidity is optimized.

Introduction

Selective Laser Melting (SLM, also referred to as LBM: Laser Beam Melting) is the most common process for Additive Manufacturing (AM) of metal parts in modern industrial machines. Current performances of AM machines enabled AM processes for production of small series with the attendant technological challenges constantly growing in both industry and academia [1]. The performance of the SLM process chain hinges on five major points of interest [2]: ‘Equipment’, ‘Material’, ‘Production’, ’Batch’, ‘Part Finishing’. Certain physical aspects of metal powder melting remain lesser known, and a current research trend is oriented towards modelling these physical phenomena [3]. The work presented in this paper deals with the mechanical actuators control of the production machine.

In the ‘Equipment’ category, the main actuator element for SLM machines is the three-axis deflection and focus control system (Fig. 1). The control strategy of the galvanometer motor which drives each axis of the actuator is addressed here.

The galvanometer motor is mostly used in a closed-loop control framework [4]. The controller structure is synthetized so as to meet the precision and rapidity requirements of the operation. Control strategies of the galvanometer motor have been extensively studied in the automatic control field, as an example of high precision system. Reference [5] proposes a general framework in which the synthesis of the controller can be divided into four main points of interest towards the improvement of the closed-loop dynamics: ‘Feedforward/ Command Shaping’, ‘Feedback Control’, ‘Modeling/Identification’ and ‘Optimization/ Auto tuning’. The most-encountered conventional structures in industry are Proportional-Integral-Derivative (PID)-based controllers [4]. State of the art works supply a large panel of advanced controller schemes that tackle specific problems such as residual vibration suppression by initial value compensation techniques [6], disturbance rejection through enhancement of the controller with an observer [7] or online closed-loop dynamics enhancement with adaptive control strategies [8].

In this paper, we propose an H-infinity robust controller [9] structure synthesis based on the desired closed-loop behavior of the galvanometer system. Compared to the dynamics enhancement techniques mentioned above, the H-infinity controller synthesis takes into account multiple synthesis objectives all at once (stability, rapidity, precision and torsional bending modes rejection). Moreover, H-infinity robust controllers have already been extensively used in industry for mechanical servo systems [10]. Performance achieved with actuator axes driven with PID-based conventional structures and the newly introduced H-infinity structures are compared in this work within the framework of an experimental marking job. It is shown that the operation precision is improved with H-infinity structures driving the actuator axes when operation rapidity is optimized. An open architecture test bench is developed with a conventional control structure, following the H-infinity synthesis technique is applied to the control of galvanometers.

Section snippets

The open-architecture test bench

Experiments are carried out on a 2-axis open architecture test bench as presented in Fig. 2. The setup, designed to reproduce the real behavior of a commercial system, consists of a 2-axis commercial galvanometer scanner system (motor and mirror) along with its control card. The addition of an external control card allows for an open-architecture framework, providing implementation of proposed control structures. Developed control strategies are implemented in Matlab/Simulink® in a PC

H-infinity controller synthesis

The main benefit of H-infinity controllers is the possibility to keep precision indicators degradation as low as possible while increasing system rapidity. As a matter of fact, H-infinity controller synthesis techniques allow simultaneous consideration of multiple control objectives which impact stability, rapidity and precision of the system.

H-infinity controller synthesis strategy [9] consists of the feedback controller Ks synthesis for a given system plant Ps, as depicted in Fig. 6 as the

Results on a marking job with optimized scanning speed on the 2-axis experimental test bench

A SLM marking job is performed with different operation speed settings on the open architecture test bench in the case of axes driven by the conventional and the H-infinity control structures. The job consists in the realization of 100 cubes whose side length is equal to 10 mm as illustrated in Fig. 13.

Results are given for the realization of a single cross-section (layer) of the job in the x,y plane, the z dimension can be taken into account with the value of the layer thickness of the SLM

Conclusions

This paper presented the achievable benefits of an H-infinity controller structure for the control of the galvanometer motor, which is the main actuator of SLM machines for AM. In comparison to conventional PID-based control, results obtained on an open architecture experimental test bench show that this advanced controller increases the system precision performance while operation speed is optimized. Additive manufacturing ‘Key Performance Indicators’ (KPI) such as part porosity and melt pool

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work takes part in the SOFIA project context. SOFIA is an applied research program for metal additive manufacturing, initiated by AddUp, a Fives & Michelin Joint Venture dedicated to Additive manufacturing solutions. SOFIA is funded by the Auvergne Rhone-Alpes region as well as Bpifrance as a structuring R&D project for competitiveness, within the’ Investing for the Future’ program.

References (17)

There are more references available in the full text version of this article.

Cited by (0)

View full text