Fig. 1. In-situ monitoring system for optical signal of the SLM melt pool. (a) Schematic of monitoring device; (b) physical picture of coaxial monitoring device

Fig. 2. Size and scanning strategy of tensile test pieces. (a) Dimension diagram of tensile test pieces; (b) schematic of blocks and tensile test pieces scanning strategy

Fig. 3. Signals collected without laser operation. (a) With illumination and without filtering; (b) without illumination and without filtering; (c) with illumination and with filtering; (d) without illumination and with filtering

Fig. 4. Dynamic characteristics of melt pool signal. (a) Before filtering; (b) after filtering

Fig. 5. Changes before and after removing jump signals. (a) Original signal; (b) effective signals after removing jump signals by direct method; (c) radiation signal in single-pass scanning; (d) effective signals after removing jump signals by segmented method; (e) frequency spectrum without removing jump signals; (f) time-domain waveform of a single scanning-line

Fig. 6. Surface morphology of formed samples at different process parameters. (a) Surface morphology of formed samples at different laser powers; (b) surface morphology of formed samples at different scanning speeds; (c) surface morphology of formed samples at different scanning spacings; (d)‒(h) local magnified images

Fig. 7. Variation of melt pool radiation light signal intensity at different process parameters. (a) At different laser powers; (b) at different scanning speeds; (c) at different scanning spacings

Fig. 8. Correlation between optical signals intensity mean and surface roughness

Fig. 9. Relationship between SLM-formed sample density and time-domain signal intensity mean. (a) At different laser powers and scanning speeds; (b) at different scanning spacings

Fig. 10. Correlation between mechanical properties of the formed sample and optical signal intensity mean. (a) Correlation between tensile strength and optical signal intensity mean; (b) correlation between yield strength and optical signal intensity mean; (c) correlation between elongation at break and optical signal intensity mean; (d) correlation between elastic modulus and optical signal intensity mean

Fig. 11. Internal defects distribution of the samples formed at different parameters. (a)‒(d) At different laser powers; (e)‒(h) at different scanning speeds; (i)‒(m) at different scanning spacings

Fig. 12. Correlation between optical signal intensity mean and porosity. (a) At different laser powers; (b) at different scanning speeds; (c) at different scanning spacings

Fig. 13. Time-series signal and frequency spectrum of the sample formed at process 1. (a) Time-series signal; (b) frequency spectrum

Fig. 14. Characteristics of the optical signals. (a)(b) Frequency-time distribution of melt pool signals obtained by wavelet analysis after removing jump signals; (c) characteristics of radiation signals at laser power of 20 W; (d) characteristics of radiation signals at laser power of 40 W; (e) characteristics of radiation signals at laser power of 60 W; (f) characteristics of radiation signals at scanning speed of 1000 mm/s; (g) characteristics of radiation signals at scanning speed of 500 mm/s; (h) characteristics of radiation signals at scanning speed of 100 mm/s

Fig. 15. Three-level wavelet packet decomposition

Fig. 16. Frequency spectrum and envelope spectra of radiation signals under different processes. (a) Frequency spectrum under process 1; (b) envelope spectrum under process 1; (c) envelope spectrum under process 2; (d) envelope spectrum under process 3; (e) envelope spectrum under process 4