Laser triangulation for coaxial height measurement in laser metal deposition

Share this on social media:

Issue: 

Ali Gökhan Demir, of the Politecnico di Milano, reports on the use of laser triangulation to address a significant challenge being faced by laser metal deposition users

This article was co-authored by Simone Donadello and Barbara Previtali of the Politecnico di Milano 

Metal additive manufacturing (MAM) has gained an important pace during the last few years. A key contributing factor to this progress is the role of lasers in MAM processes. Indeed, the last two decades have seen the rise of industrial high-power solid state lasers – namely, fibre, disc and diode lasers. The robust architecture of these systems, the flexible control of the optoelectronic device, the high beam quality – as well as an improved availability of service and reduced costs – have been some of the important advances that lead to such industrial acceptance. 

Concerning MAM processes, a highly flexible and robust digitally-controlled energetic beam appears a natural match. Most laser powder bed fusion (LPBF) systems operate with high-brilliance fibre lasers for process durations measured by days. On the other hand, market share of directed energy deposition (DED) processes is limited. 

Laser metal deposition (LMD), which unites the laser beam with blown powder feedstock, is arguably the most industrially mature DED process. LMD could be seen as a natural derivation of the laser cladding process, which has long been used in industry. In the most common form, the process uses powder feedstock blown from coaxial or multijet nozzles encased in a processing head that projects the laser beam to the deposition area. Several industrial solutions are present where cartesian axes and robotic systems are employed for beam positioning and workpiece handling.

Inline height measurements required in LMD

Similar to other MAM processes, LMD suffers from a lack of design rules, industrial standards and stable process parameters for new materials. In addition, process robustness and dimensional accuracy can be compromised, especially in long depositions.

One of the biggest differences of LMD compared to the other MAM processes is related to the build rate variability. Such variability in long process runs manifests a height mismatch between the commanded height increment of the deposition head and the deposited layer thickness. The workpiece can grow faster or slower than expected, changing the standoff distance between the nozzle and the deposition region. Consequently, the workpiece geometry can be wrong, or in some cases the process can fail completely. The height mismatch can be related mainly to heat accumulation and geometrical changes, whereas issues related to the nozzle or the powder feeder can also be the underlying cause.

In scientific literature, a self-regulation phenomenon has been reported, where a stable standoff distance is reached over a certain time1. However, this self-regulation of the standoff distance is less likely when complex parts with variable sections are manufactured with variable sections and inclination angles – see Figure 1. In industrial practice, human operators often intervene with the process parameters when such geometrical derivations are observed. This requirement of a skilled operator to maintain process stability over prolonged process durations is not a feasible option, however.

Figure 1: Examples of height mismatch errors leading to part failiure in complex trajectories. (Image: Demir et al.)

Limits of existing methods

The measurement of the workpiece dimensions in LMD has been an object of scientific research as well as industrial practice2. Several options based on tactile measurements, projection techniques, as well as tomographical techniques have been tested and employed. While these techniques have particular strengths in terms of the temporal or spatial resolutions provided, they also have disadvantages. Tactile measurements, projection techniques and x-ray tomography commonly require the LMD process to stop in order for the measurement to be taken. Interferometric measurements can be employed online and coaxially, but these require expensive equipment. Optical pyrometers can be used to provide an indirect measurement through the average temperature of the melt pool area. 

A novel solution

Laser triangulation is a well-known method for dimensional measurements which has been employed in LMD3, as well as in 3D scanners and camera vision systems – however, their typical off-axis configuration limits their applicability to real-world cases. 

At the Politecnico di Milano, the triangulation method has been developed and demonstrated on an existing Additube LMD system from BLM Group4. The triangulation measurement has been implemented in an innovative coaxial configuration, allowing for an omnidirectional height measurements with big advantages in terms of flexibility and robustness. Such implementation provides an important tool for real-time process monitoring. 

Figure 2 shows the schematic functioning principle of the coaxial laser triangulation system launched into an industrial LMD head (MWO-I, Kuka Reis) mounted on a six-degrees of freedom industrial robot (IRB 4600-45, ABB) and a two-axis rotary table, employing a 3kW multimode active fibre laser beam source (YLP-3000, IPG Photonics). The probe beam shares the process laser optical path, and it is tilted with an angle with respect to the normal of the substrate. The position of the probe beam is viewed through a coaxial commercial digital camera, and consequently the lateral displacement viewed by the camera is converted to the height information. The current system runs with a height resolution in the order of 0.1mm and an acquisition frequency higher than 100Hz, suitable for most industrial LMD applications.

Figure 2: A schematic of the coaxial laser triangulation functioning principle (left) and the integration on an Additube LMD system at the Politecnico di Milano (right) 

Advantages of coaxial triangulation

The implemented coaxial triangulation system has been tested with several different geometries. Figure 3 shows a case study where a cylindrical form is produced employing AISI 316 stainless steel powder. 

The robotic arm is programmed for a constant height increment superimposed to a circular motion, resulting in a single helical trajectory. With a 35mm diameter and wall thickness of 1.5mm, such form is representative of geometries prone to heat accumulation and thus height mismatch.

Extracting the position data from the robot controller, the height measurement is used to generate the height error map along the deposition trajectory. The results showed that the final dimensional error can reach up to 5mm. The process settles to a constant error after the deposition of approximately 50 layers. It can be seen that the height information is highly correlated to the deposition width, since the real texture formation along the build direction can be viewed at the measured height colour map.

Figure 3: A cylindrical thin-walled structure built by LMD and the corresdonding measured height error as a function of time. The colour map shows the height error profile of a cyindrical sample. Positive height error refers to a decrease in standoff distance.

Another critical condition is related to the deposition of thin-walled structures. Such geometries are prone to height variations, also due to the change in trajectory length, as well as the dynamics of the mechanical axes. Figure 4 shows a case study where a stepped thin-wall profile was produced using AISI 316 stainless steel powder. It can be seen that the height error map follows the shape of the real part with high fidelity, including the acceleration zones at the end of the trajectory. The coaxial triangulation measurements were compared to measurements taken with a touch probe after the deposition. The results indicated in Figure 4 show the efficacy of coaxial triangulation as a contactless and inline measurement tool.

The coaxial laser triangulation is based on a simple principle, which becomes effective through correct implementation of hardware and software solutions. The strength of the method derives from the fact it provides a direct measurement of a fundamental process output through a non-invasive optical principle, rather than an indirect measurement which requires calibration for different configurations and materials. It is highly suitable for retrofitting on existing systems with monitoring portals available on the process head.

Figure 4: The actual photograph and height error profile of a thin-walled sample. Positive height error refers to a decrease in standoff distance.

The system can be used for online measurement of the workpiece height as a means for process monitoring, as well as for building control actions with the analysed signal. The output can be used as an aid for 3D reconstruction of the workpiece, hence part qualification. The measurement principle can allow for a real autonomous operation of the LMD systems with prolonged process duration, opening up new application fields. The development of such solutions, as well 
as industrialisation of the product, are ongoing.

The authors acknowledge the collaboration with BLM Group. This work was supported by European Union, Repubblica Italiana, Regione Lombardia and FESR for the project MADE4LO under the call ‘POR FESR 2014–2020 ASSE I –AZIONE I.1.B.1.3‘.

References

[1] G. Zhu, D. Li, A. Zhang, G. Pi, Y. Tang, The influence of laser and powder defocusing characteristics on the surface quality in laser direct metal deposition, Opt. Laser Technol. 44 (2012) 349–356. doi:10.1016/j.optlastec.2011.07.013.

[2] S.K. Everton, M. Hirsch, P. Stravroulakis, R.K. Leach, A.T. Clare, Review of in-situ process monitoring and in-situ metrology for metal additive manufacturing, Mater. Des. 95 (2016) 431–445. doi:10.1016/j.matdes.2016.01.099.

[3] A. Heralić, A.K. Christiansson, M. Ottosson, B. Lennartson, Increased stability in laser metal wire deposition through feedback from optical measurements, Opt. Lasers Eng. 48 (2010) 478–485. doi:10.1016/j.optlaseng.2009.08.012.

[4] S. Donadello, M. Motta, A.G. Demir, B. Previtali, Monitoring of laser metal deposition height by means of coaxial laser triangulation, Opt. Lasers Eng. 112 (2019) 136–144. doi:10.1016/j.optlaseng.2018.09.012.

Navigation

Navigation

Navigation

Navigation

Navigation

Navigation