Optical emission spectroscopy of tungsten inert gas wire-arc plasma for in situ monitoring in additive manufacturing

Abstract

In situ diagnostics for defect detection and compositional assessment remain critical for addressing persistent challenges in Additive Manufacturing (AM). Among the available diagnostic techniques, Optical Emission Spectroscopy (OES) stands out for its strong capability to provide in situ and non-intrusive monitoring of the AM process. As part of our ongoing efforts to develop a low-cost, real-time AM monitoring tool, this study represents an initial investigation focused on the concentration of key species in the Tungsten Inert Gas (TIG) wire-arc plasma generated during welding of an aluminum (Al-6061-T6) sheet using OES. Emission spectra from the arc plasma were analysed to identify and quantify key spectral lines associated with aluminum and potential alloying elements. The plasma temperature and electron number density were estimated to support quantitative analysis through Boltzmann and Saha methods to enable accurate concentration determination by calibration-free approach. The results demonstrated the potential of OES as an effective tool for monitoring and optimizing the aluminum welding operations through in situ non-intrusive plasma composition analysis.

Keywords

tungsten inert gas arc plasma aluminum optical emission spectroscopy calibration-free optical emission spectroscopy plasma temperature electron number density local thermodynamic equilibrium condition and concentration

Introduction

Additive manufacturing (AM) has emerged as a transformative technology across diverse industrial sectors, including aerospace, biomedical, and automotive engineering.^1–4 Among Contemporary AM process, wire arc AM (WAAM)-which employs an electric arc as the heat source (tungsten inert gas (TIG) or MIG) in conjunction with a metallic wire feedstock has gained significant attention.^1,4 However, defects, including composition loss, porosity, cracks, and low repeatability often render components unsuitable for many practical applications.⁵ Therefore, monitoring and controlling the welding process is essential to ensure build quality. Both traditional methods, such as measuring welding current, arc voltage, gas flow rate, welding speed and wire speed rate⁶; and non-traditional approaches, including optical,^1,7–9 acoustic,^1,10 X-ray,^8,11,12 Ultrasonic^13,14 and thermal sensing,¹⁵ have been implemented to monitor and control the welding processes, ensuring consistent build quality.^6,16,17 However, most of these techniques are often limited by inherent constraints, higher costs, resource and time-intensive.^10–12,15 Consequently, cost-effective monitoring methods using arc plasma optical emissions have gained attention. In WAAM, arc plasm is the highly energetic, ionized region between the cathode and workpiece, which governs the energy transfer and weld pool formation. While the cathode is heated via Joule heating and collisions with ions and atoms, electron emission provides cooling, and near-surface plasma reactions, including ionization and recombination, influence particle dynamics. The emitted arc light contains critical information on plasma temperature, composition,^{1,7,8,14,18–28} which in turn are strongly correlated with welding parameters and weld quality. Optical emission spectroscopy (OES) was relatively a simple but powerful technique that has been widely used for the diagnosis of arc plasma to measure plasma temperature and electron number density,^{18,20,25,26,29,30} non-intrusive monitoring of welding process,^19,24,31 anode metal vapors,^22,32 and the shielding gases.^21,25,27 More recently, OES integrated with artificial intelligence enhanced defect detection and closed-loop process control during in situ monitoring of AM processes.^1,14,33

Aluminum, though one of the most economical metals for its corrosion resistance, high strength-to-weight ratio, and alloying flexibility, remains challenging to process by AM. Alloys like 2xxx, 5xxx, and 6xxx, common in automotive applications, are prone to hot cracking, porosity, and compositional variations during fabrication.^5,11,34–36 These challenges are further exacerbated by aluminum's high reflectivity and thermal conductivity, strong tendency for porosity formation, and the rapid solidification rates inherent to AM processes.^{5,11,31,35,36} Additionally, the weld metal composition can be significantly altered due to the selective vaporization of volatile alloying elements, particularly magnesium, thereby adversely affecting mechanical properties, corrosion resistance, and hot crack susceptibility.^31,37,38 Cieslak et al,³⁷ reported that magnesium vaporization during laser welding leads to a measurable decrease in weld metal hardness compared to the base material, with alloys such as 5456 and 5086 exhibiting pronounced softening due to reduced magnesium content in the weld pool. Furthermore, compositional and microstructural heterogeneities across the weldment can degrade corrosion resistance. Zhang et al³¹ proposed the correlation between the spectral parameters and structural parameters such as forming width, composition, grain size, and porosity defects. Also, Aluminum alloy susceptibility to hot cracking is strongly composition-dependent.^11,34–38 Therefore, making real-time compositional assessment, alongside plasma temperature and properties, essential for effective in situ welding monitoring and control.^{1,14,34,37,39} However, continuous monitoring of metal composition during welding typically requires multiple standard reference materials to establish and maintain calibration curves, which can be time-consuming and impractical in dynamic manufacturing environments.³⁹ To address this limitation, Calibration-free OES (CF-OES) emerged as a promising alternative by using plasma parameters along with spectroscopic constants to determine species concentrations without external standards. This reduces experimental complexity and cost while enabling rapid, real-time compositional analysis, ideal for real-time monitoring.

As part of our ongoing efforts to develop a low-cost, real-time WAAM compositional monitoring tool based on CF-OES, this study presents an initial investigation focused on the OES analysis of TIG arc plasma generated from aluminum (Al 6061-T6) sheets, aiming to characterize key plasma parameters, including plasma temperature, electron number density, and species concentration to assess their relevance for quality control monitoring in WAAM. By analysing the characteristic emission lines from aluminum plasma, as well as shielding gas (argon) constituents, species concentration values were obtained using CF-OES under varying arc current conditions.

Experimental setup

The experimental setup was designed to perform in situ OES of TIG wire-arc plasma generated during the welding of commercially available aluminum sheet (Al 6061-T6, M/s McMaster-Carr) measuring 6″ × 12″ with a thickness of 1.0″ as shown in Figure 1. A Miller Dynasty 350 Dx constant-current TIG welding power source was used with a non-consumable tungsten electrode and pure argon shielding gas at a flow rate of 15 L/min for the arc plasma generation as shown in Figure 1(a).

Figure 1.

Experimental setup for OES collection from TIG arc Al plasma. OES: optical emission spectroscopy; TIG: tungsten inert gas.

Optical emissions from arc plasma were collected at 45° angle above the base plate perpendicularly to welding direction using a simple 2-m long UV–Visible Optical Fiber (QP400-1-UV-VIS, M/s Ocean Optics) of 400 µm core diameter with 0.22 NA from 2-feet distance thus avoiding the intensity attenuation setup to observe the total emissions from the Al arc plasma without saturating the spectrometer as shown in Figure 1(b). The collected optical light was delivered to a low-cost, low-resolution (∼ 1.5 nm) wavelength and Intensity calibrated handheld spectrometer (FLAME-T-UV-VIS-mini, M/s Ocean Optics) equipped with 600 lines/mm grating blazed at 300 nm and covering a wavelength range of 185–1047 nm. The collected spectra were time-integrated over 10 ms intervals to capture steady-state plasma behavior. Such 10 spectra were collected to ensure reproducibility and calculate the plasma parameters with error bars. The optical spectrum was collected at three different arc currents: 150, 200 and 250 A to visualize the effect of arc power on the plasma parameters. The ablated regions of the Al sheet were shown in Figure 1(c). Due to the intense visible and ultraviolet radiation emitted by the TIG arc, stringent eye protection measures were implemented while collecting the optical spectra, Operators wore welding helmets with auto-darkening filters rated to shade level 10 or higher for eye protection and the spectrometer was handled from a safe place to collect the data.

Results and discussion

Analysis of OES spectrum

The averaged optical spectrum acquired from TIG arc plasma of Al (6061-T6) sheet was shown in Figure 2. The spectra showed strong atomic emissions of Aluminum (Al I) at 309.697, 394.44 and 396.15 nm, Magnesium ionic (Mg II) and atomic (Mg I) emissions at 280.2, 285.6, 294.2 and 383.63 nm, respectively. Trace emissions of Iron (Fe II), Titanium (Ti II), Chromium (Cr I), Manganese (Mn II), Nickel (Ni I), Calcium (Ca II) and Copper (Cu I) were also observed, indicating minor contamination or alloying additives. Apart from these, the spectrum also showed several strong emissions from the argon shield gas (Ar I) in 660–860 nm wavelength range compared to Al emission lines intensity. Argon was a monoatomic gas, so it does not lose energy through vibration/rotational modes, and the high flow rate of Ar gas could increase the Ar gas density near the arc by pushing the quickly ionized and condensed Al vapor away from the hotter and stable arc plasma core. Thus, more energy was available for Ar electronic excitations due to its relatively low excitation threshold (∼11.5 eV) at arc center led to the stronger Ar optical emissions in 660–860 nm range than Al metal and its impurities. The standard wavelengths and their corresponding spectroscopic parameters of each emission line were summarized in Table 1 with the help of NIST atomic database.⁴⁰ All emission spectrum line intensities were prone to fluctuations due to the probabilistic behavior of the plasma breakdown process, fluctuations in the TIG plasma source and light transmission through the optical elements, etc. To maintain a better signal-to-noise ratio and evaluate the experimental uncertainties in the plasma properties, the experiment was repeated 10 times at the same experimental conditions.

Figure 2.

Averaged optical spectrum of tungsten inert gas (TIG) arc Al plasma at 150, 200 and 250 A arc currents.

Table 1.

Spectroscopic Parameters of Tungsten Inert Gas (TIG) Arc Plasma of Al (6061-T6) Sheet.

Species	NIST λ (nm)	E_k (eV)	E_i (eV)	A (s^-1)	g_k	g_i
Al I	309.283	4.0214836	0.0138938	1.16 × 10⁰⁷	4	4
	394.400	3.1427212	0	4.99 × 10⁰⁷	2	2
	396.152	3.1427212	0.0138938	9.85 × 10⁰⁷	2	4
Fe II	225.182	10.2725501	4.76830747	4.40 × 10⁰⁸	12	10
Fe II	249.326	7.6061288	2.63486052	3.04 × 10⁰⁸	16	14
Cu II	237.630	14.34065139	9.12471804	1.61 × 10⁰⁸	3	3
	240.011	8.42057990	3.25638920	6.90 × 10⁰⁶	3	5
	642.388	17.24861095	15.31909404	5.60 × 10⁰⁷	7	5
Mn II	257.610	4.8114165	0	2.80 × 10⁰⁸	9	7
Mn II	261.020	8.163991	3.415425	3.00 × 10⁰⁸	15	13
Mg II	280.270	4.422431	0	2.57 × 10⁰⁸	2	2
Mg I	285.212	4.3458029	0	4.91 × 10⁰⁸	3	1
	294.199	6.929700	2.7166398	6.83 × 10⁰⁶	3	5
	383.829	5.9459158	2.7166398	1.61 × 10⁰⁸	7	5
	518.360	5.1078270	2.7166398	5.61 × 10⁰⁷	3	5
Ca II	317.933	7.049551	3.150984	3.60 × 10⁰⁸	4	4
Ti II	330.880	3.88083749	0.13481497	4.36 × 10⁰⁶	10	8
Ti II	334.187	4.28283327	0.57388034	1.68 × 10⁰⁸	8	6
	338.375	3.66304545	0	1.39 × 10⁰⁸	6	4
Cr I	416.551	7.4246850	4.4490806	7.50 × 10⁰⁷	13	11
	420.446	6.9275970	3.9795577	3.10 × 10⁰⁷	11	13
	426.995	5.98972002	3.08689206	3.20 × 10⁰⁶	11	9
	427.481	2.89953685	0	3.07 × 10⁰⁷	7	7
	433.944	3.83921781	0.98287742	6.92 × 10⁰⁶	7	5
	435.176	3.87826474	1.03000806	1.20 × 10⁰⁷	11	9
	443.216	5.66846045	2.87187333	1.80 × 10⁰⁷	3	1
Ni I	603.929	6.2880937	4.2357049	5.70 × 10⁰⁵	3	5
Ni I	605.368	6.2832160	4.2357049	2.20 × 10⁰⁶	7	5
H I	656.279	12.0875052	10.1988358	4.4101 × 10⁰⁷	18	8
Ar I	675.283	14.74254085	12.90701530	1.93 × 10⁰⁶	5	3
	687.128	14.71089810	12.90701530	2.78 × 10⁰⁶	3	3
	696.543	13.32785705	11.54835442	6.40 × 10⁰⁶	3	5
	706.721	13.30222747	11.54835442	3.80 × 10⁰⁶	5	5
	727.293	13.32785705	11.62359272	1.83 × 10⁰⁶	3	3
	738.398	13.30222747	11.62359272	8.5 × 10⁰⁶	5	3
	751.465	13.27303810	11.62359272	4.00 × 10⁰⁷	1	3
	763.510	13.17177770	11.54835442	2.45 × 10⁰⁷	5	5
	772.420	13.32785705	11.72316039	1.17 × 10⁰⁷	3	1
	794.817	13.28263902	11.72316039	1.86 × 10⁰⁷	3	1
	801.478	13.09487256	11.54835442	9.30 × 10⁰⁶	5	5
	811.531	13.07571571	11.54835442	3.30 × 10⁰⁷	7	5
	826.452	13.32785705	11.82807116	1.53 × 10⁰⁷	3	3
	840.821	13.30222747	11.82807116	2.23 × 10⁰⁷	5	3
	842.464	13.09487256	11.62359272	2.15 × 10⁰⁷	5	3
	852.144	13.28263902	11.82807116	1.39 × 10⁰⁷	3	3

The spectrum was collected at different TIG arc currents: 100, 150, and 200 A as shown in Figure 2 to understand the effect of arc power on the emission line intensity and plasma properties especially, the plasma temperature, which plays a key role in measuring the species composition. As the arc current increased the arc plasma power also increased, led to increased intensity of the optical emissions as shown in the collected spectrum. For better visualization of the variation of each emission element intensity with increased arc current, the collected spectrum was divided into multiple wavelength region spectra with identification of the important optical element emission wavelength from the TIG arc plasma of Al sheet was showed in Figure 3(a-l).

Figure 3.

Different optical emission lines and their intensities observed in tungsten inert gas (TIG) arc Al plasma.

The increment in the arc current increased all the optically emitted element intensity except the Al emissions as shown in Figure 3(d) and (g). As the Al has low boiling point (∼ 2743 K), the metal vaporization from Al is modest at low currents or short arc times. So, at higher arc currents the Aluminum species (Al I, Al II) can self-absorb at certain transitions due to the optical thickness of the Al plasma, reducing the net emission intensity from the metal itself. The self-absorption was an unavoidable phenomenon at higher plasma temperatures, such as arc plasma, which causes the observed line intensity to be lower than expected and elemental line shape distortions lead to inaccurate temperature calculations and misleading spectral ratios, thus causing an underestimation of element concentration making to think the species is less abundant in the test sample.

Plasma temperature (T_e) and electron number density (n_e)

Plasma temperature is a critical thermodynamic property that governs various plasma characteristics, including the molecular distribution across energy levels, the relative abundance of different ionization states, optical thinness, electron number density, species concentration and the overall behavior of the plasma. Two primary methods are prevalent in measuring plasma temperature: line intensity ratio and the Boltzmann plot method.^41,42 Although the Boltzmann plot method stands out as the best preferred approach, in the case of self-absorption prone plasmas such as TIG arc plasmas, the line intensity ratio method gives a better plasma temperature estimation. The line intensity ratio method leverages the analysis of multiple spectral lines emitted by the same element in the same ionization state. With the assumption of local thermodynamic equilibrium (LTE), the plasma temperature can be calculated from the relative intensity ratio of spectral lines from the same element ionization stage according to the Boltzmann distribution, using the equation,^41,42

\frac{I_{1}}{I_{2}} = (\frac{g_{1} A_{1}}{g_{2} A_{2}}) (\frac{λ_{2}}{λ_{1}}) \exp (\frac{- (E_{1} - E_{2})}{k_{B} T_{e}})

(1)

where I_i and λ _i are the integrated line intensity and wavelength, respectively. A_i, g_i, and E_i are the transition probability, statistical weight, and excited energy for the upper level, respectively.

k_{B}

and T_e are the Boltzmann constant and the electron temperature of the plasma, respectively. Accurate determination of plasma temperature using the above methods relies on careful selection of emission lines. Choosing distinct lines well-separated from neighboring lines minimize overlapping spectral features and ensures reliable intensity measurements. Furthermore, integrating the area under the spectral envelope provides a more accurate representation of the total emission intensity than solely relying on peak height measurements, which can be susceptible to noise or baseline fluctuations. Different elements (Al, Mg, Ti, Cr, and Ar) emission lines were used to estimate the plasma temperature, but most of the plasma temperatures resulted in errors, except the Mg emission lines. Therefore, in this study, baseline-corrected emission lines of Mg II at 280.27 nm and Mg I at 285.212, 294.199, 383.829 and 518.36 nm, along with their corresponding spectroscopic parameters from Table 1, were substituted into Equation (1) to estimate the plasma temperatures using the line intensity ratio method. Among these, the line intensity ratios of Mg II (280.27 nm)/Mg I (518.36 nm), Mg I (285.212 nm)/Mg I (294.199 nm) and Mg I (285.212 nm)/Mg I (383.829 nm) yielded reliable and accurate plasma temperatures. In general, the spectral line intensities and consequently the plasma temperatures were expected to exhibit a clear increasing trend with the arc current however, only the line ratio of Mg II (280.27 nm)/Mg I (518.36 nm) had shown a linear, monotonic progression with increasing arc current; whereas the remaining ratios show the opposite or non-monotonic behavior (see supplementary material). So, the plasma temperatures estimated using the Mg II (280.27 nm)/Mg I (518.36 nm) line ratio were 0.387 ± 0.05, 0.495 ± 0.09 and 0.557 ± 0.06 eV at 150, 200 and 250 A arc currents as shown in Figure 4(a). The estimated plasma temperatures were very low compared to the arc core temperature values mentioned in the literature,^18,20,25,27 suggesting that the spectrum was collected from the outer region of arc plasma known as cold plasma region due to our spectra collection distance. This geometric effect artificially reduced the observed plasma temperature, even if the core plasma temperature has not decreased significantly.

Figure 4.

Plasma temperature (T_e) and electron number density (n_e) for 150, 200 and 250 A arc currents.

The electron number density (n_e) of the plasma was estimated from the well-resolved Mg neutral and singly ionized lines at 280.27 and 285.21 nm using Saha-Boltzmann equation method. The Saha-Boltzmann equation was given by^41,42:

n_{e} (in m^{- 3}) = \frac{2 {(2 π m_{e} k T)}^{\frac{3}{2}}}{h^{3}} \frac{I_{m n}^{I} g_{i}^{I I} A_{i j}^{I I}}{I_{i j}^{I I} g_{m}^{I} A_{m n}^{I}} e^{\frac{- (E_{ion} + E_{i}^{I I} - E_{m}^{I})}{k T}}

(2)

where I is the line intensity; g, A and E were statistical weights, Transition probability and Energy level values of Mg neutral and ionic lines, respectively. E_ion is the first ionization potential of the Mg, m_e is the electron mass, k is the Boltzmann constant, h is the Planck constant and T is the plasma temperature. The estimated electron number densities were 1.223 ± 0.44 × 10¹¹, 1.752 ± 0.23 × 10¹³ and 1.466 ± 0.87 × 10¹⁴ cm⁻³ at 150, 200 and 250 A arc currents, respectively as shown in Figure 4(b). The electron number density shown in the non-linear monotonic progression with increasing arc current.

Local thermal equilibrium condition and McWhirter criteria

The assumption of LTE was a fundamental for OES-based diagnostics of plasmas. Under LTE, the populations of atomic and ionic excited states are governed by the Boltzmann distribution, while the balance between different ionization stages follows the Saha equation, both characterized by a single plasma temperature. This condition is valid when collisional processes dominate over radiative processes, ensuring that the excitation and ionization kinetics are controlled by local electron collisions rather than photon exchange.^41,42

According to McWhirter criteria, the necessary but insufficient condition for LTE was given by

n_{e} ({cm}^{- 3}) \geq 1.6 \times 10^{12} T^{\frac{1}{2}} (Δ E)^{3}

(3)

where T is the plasma temperature (in K) and ΔE is the excitation energy gap (in eV) from the ground state to the lowest level. For TIG arc plasma of Al sheet, the McWhirter criteria was estimated as: 1.98 $\times 10^{16}$ cm⁻³, which is higher than the number density calculated at three plasma currents 150, 200, and 250 A, respectively as shown in Figure 4(b). This indicates that the LTE condition was not satisfied by the TIG arc plasma.^26,43 However, several studies showed that the temperature and electron number densities at the arc plasma core did satisfy the LTE condition.^{24–26,31,43} Therefore, this discrepancy may be attributed to the inaccurate plasma temperature estimation due to the collection of optical spectra from the peripheral regions of the arc (as the collection distance was 2 feet) and from the cooling plasma plume above the Al plate, where the electron density drops below 10¹⁶ cm⁻³ and strong radiative losses occur.¹⁸ So, the species concentration was estimated using Boltzmann and Saha distributions under partial LTE conditions.

Species concentration measurement

In the typical time window of the OES experiment, the elemental concentration in the plasma was represented by neutral atoms and singly charged ions of the corresponding element to a significant degree. Hence, the concentration of an element (C_s) will be proportional to the sum of population number densities (n^s) of neutral atoms and singly charged ions^41,42 i.e.,

C_{s} = n_{neutral}^{s} + n_{ions}^{s}

Under partial Local Thermal Equilibrium (LTE) conditions, the population number densities (n^s) of Al, Mg and Ti lines were estimated using Boltzmann distribution in terms of Integrated line intensity^41,42:

n^{s} = \frac{I_{i j} U_{s} (T)}{A_{i j} g_{i}} e^{\frac{E_{i}}{k T}}

(3)

where I is the intensity or area of the corresponding line, U_s(T) is the partition function; g, A, and E were statistical weights, Transition probability and Upper Energy level (in eV) values of corresponding lines respectively, T is the plasma temperature (in K) and k is the Boltzmann constant (in eV/K). From the observed OES spectrum, we have only Al I, Mg I, Mg II and Ti II elemental lines in the spectrum, so the missing Al II and Ti I population number densities were estimated using the Saha equation^41,42:

n_{e} = \frac{n^{I}}{n^{I I}} \frac{{(2 π m_{e} k T)}^{3 / 2}}{h^{3}} \frac{2 U_{s}^{I I} (T)}{U_{s}^{I} (T)} e^{- E_{ion} / k T}

(4)

where, n^I is the neutral population number density of element, n^II is the ionic population number density of element, n_e is the electron number density (in m⁻³), E_ion is the ionization energy of element (in eV);

U_{s}^{I I} (T)

and

U_{s}^{I} (T)

is the ionic and neutral partition functions, T is the plasma temperature (in K) and k is the Boltzmann constant (in eV/K). The estimated neutral and ionic population number densities of Al, Mg and Ti lines were used to estimate the species concentration using equation (4) in mol.% and they can be converted into wt.% using below equations:

C_{s} (mol . %) = \frac{n_{s}^{I} + n_{s}^{I I}}{\sum_{s} n_{s}^{I} + n_{s}^{I I}} 100

(5)

C_{s} (wt . %) = \frac{C_{s} (mol . %) \times {MW}_{s}}{\sum_{s} C_{s} (mol . %) \times {MW}_{s}}

(6)

where C_s is the element concentration; $n_{s}^{I}, n_{s}^{I I}$ are the neutral and ionic population number densities of element and MW is the molecular weight of element. The estimated concentration of Al, Mg and Ti from OES spectrum TIG arc plasma of Al sheet at different arc currents are summarized in Table 2.

Table 2.

The Estimated Concentration Using Optical Emission Spectroscopy (OES) Analysis Versus Actual (Standard) Concentration of Al 6061-T6.

Element	Concentration (wt.%) (Measured)			Concentration (wt.%) (actual)
Element	150 A	200 A	250 A	Concentration (wt.%) (actual)
Al	99.96432	99.90024	99.61054	98.6
Mg	0.00875	0.03787	0.18629	1.2
Ti	0.02693	0.06188	0.20317	0.15

Error analysis

The estimated error in the concentration measurements in CF-OES from the certified values was measured using the root mean square error (RMSE) of composition as:

{RMSE}_{comp} = \sqrt[2]{\frac{1}{N} \sum_{i = 1}^{N} {(C_{i}^{CF - OES} - C_{i}^{ref})}^{2}}

(7)

where $C_{i}^{CF - OES} and C_{i}^{ref}$ are the concentration measured using CF-OES analysis and certified reference values, N is the no. of elements measured. The RMSE values estimated from the summarized concentration values (Table 2) were 1.04, 1.00 and 0.8 wt.% for 150, 200 and 250 A. Arc currents. From the results, the concentration values of Al (major species) were accurately measured using CF-OES with ± 1 wt.% at three arc currents, but for minor species (Mg and Ti) the CF-OES method was either over-estimated or under-estimated due to the self-absorption nature of the arc plasma and inaccurate plasma temperature and number density due to the spectral collection from cold plasma region.

Summary and conclusions

In summary, the OES data from the TIG arc plasma of Al sheet (Al 6061-T6) was collected using low-cost, low-resolution spectrometer from 2-feet distance without saturation thus avoiding the intensity-reducing filter setup to observe the original spectral emissions from the arc plasma. The analysis of OES data showed that the observed spectra was undergone self-absorption phenomena, commonly occurs in high-power arc plasmas. We confirmed that, for self-absorbed TIG arc plasma of Al sheet OES spectra the line intensity ratio method worked better than the Boltzmann plot method for plasma temperature calculations. Different elements (Al, Mg, Ti and Cr) emission lines were tried to estimate the plasma temperature using line intensity ratio method but only Mg lines shown in Figure 4(a) resulted in the temperature values of 4500–6500 K range, proportional to the increase of arc current. However, the estimated electron number densities with the observed plasma temperature values were below the McWhirter criteria, confirming the plasma was under partial LTE condition. This suggested that the OES spectra were collected from the outer regions of the arc plasma. The species concentrations of Al, Mg and Ti were calculated using the Boltzmann and Saha equations as explained in the results section. Although, the estimated concentrations values were either over-estimated or under-estimated due to the self-absorption and inaccurate plasma temperature values, the RMSE errors were below ± 1 wt.% for three arc currents, which were within the acceptable range. This shows that the OES analysis with low-cost, low-resolution spectrometer can be utilized for the online species concentration monitoring tool of TIG arc plasma. This will be the most cost-effective and accurate solution for at least major species concentration monitoring during the welding process.

Supplemental Material

sj-docx-1-stw-10.1177_13621718261460943 - Supplemental material for Optical emission spectroscopy of tungsten inert gas wire-arc plasma for in situ monitoring in additive manufacturing

Supplemental material, sj-docx-1-stw-10.1177_13621718261460943 for Optical emission spectroscopy of tungsten inert gas wire-arc plasma for in situ monitoring in additive manufacturing by Nagaraju Guthikonda, Goodspeed Devon, Camden Carroll, Sudarsanam S Babu, Bradely Jared and Zhili Zhang in Science and Technology of Welding and Joining

Footnotes

Acknowledgment

This research was supported by NSF and DOE.

ORCID iDs

Nagaraju Guthikonda

Camden Carroll

Sudarsanam S Babu

Bradely Jared

Zhili Zhang

Authors contributions

Nagaraju Guthikonda: Conceptualization, Investigation, Methodology, data collection, Formal analysis, and Writing-original draft and editing. Godspeed Devon: Data collection, TIG arc operation. Camden Carroll: Data collection. Sudarsanam Suresh Babu: Conceptualization and Writing-review. Jared Bradely Howell: Conceptualization, Resources, and Writing-review. Zhili Zhang: Funding acquisition, Project Administration, Resources, Supervision, and Writing-review and editing.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplemental material

Supplemental material for this article is available online.

References

Mazumder

. Design for metallic additive manufacturing machine with capability for “certify as you build”. Procedia CIRP 2015; 36: 187–192.

Tino

, et al. Additive manufacturing in radiation oncology: a review of clinical practice, emerging trends and research opportunities. Int J Extreme Manuf 2020; 2: 012003.

Eyers

, et al. The flexibility of industrial additive manufacturing systems. Int J Oper Prod Manage 2018; 28(12): 2313–2343.

Williams

, et al. Wire + arc additive manufacturing. Mater Sci Technol 2016; 32: 641–647.

Martin

, et al. 3D Printing of high-strength aluminium alloys. Nature 2017; 549: 365–369.

Ding

, et al. Wire-feed additive manufacturing of metal components: technologies, developments and future interests. Int J Adv Manuf Tech 2015; 81: 465–481.

Ogawa

. High speed imaging technique part 1 – high speed imaging of arc welding phenomena. Sci Technol Weld Joining 2011; 16: 33–43.

Tsukamoto

. High speed imaging technique part 2 – high speed imaging of power beam welding phenomena. Sci Technol Weld Joining 2011; 16: 44–55.

Chen

. Welding penetration monitoring for pulsed GTAW using visual sensor based on AAM and random forests. J Manuf Process 2021; 63: 152–162.

10.

Purtonen

Kalliosaari

Salminen

. Monitoring and adaptive control of Laser processes. Phys Procedia 2014; 56: 1218–1231.

11.

Derekar

. A review of wire arc additive manufacturing and advances in wire arc additive manufacturing of aluminium. Mater Sci Technol 2018; 34: 895–916.

12.

Martin

, et al. Dynamics of pore formation during laser powder bed fusion additive manufacturing. Nat Commun 2019; 10: 1987.

13.

Carlson

, et al. Ultrasonic NDT methods for weld sensing. Mater Eval 1992; 50: 1338–1343.

14.

Ren

Mazumder

. In situ porosity recognition for laser additive manufacturing of 7075-al alloy using plasma emission spectroscopy. Sci Rep 2020; 10: 19493.

15.

Grasso

Colosimo

. Process defects and in situ monitoring methods in metal powder bed fusion: a review. Meas Sci Technol 2017; 28: 044005.

16.

Weglowski

. Monitoring of arc welding process based on arc light emission. In: Kovacevic

(ed.) in Welding Processes. London, UK: IntechOpen, 2012.

17.

, et al. A review of the wire arc additive manufacturing of metals: properties, defects and quality improvement. J Manuf Process 2018; 35: 127–139.

18.

Valensi

, et al. Plasma diagnostics in gas metal arc welding by optical emission spectroscopy. J Phys D: Appl Phys 2010; 43: 434002.

19.

Pauna

, et al. Optical emission spectroscopy as an online analysis method in industrial electric arc furnaces. Steel Res Int 2020; 91: 2000051.

20.

, et al. Spectroscopic measurement of temperatures in pulsed TIG welding arcs. J Phys D: Appl Phys 2011; 44: 405202.

21.

Wilhelm

, et al. Study of the welding gas influence on a controlled short-arc GMAW process by optical emission spectroscopy. J Phys D: Appl Phys 2010; 43: 434004.

22.

Murphy

. The effects of metal vapour in arc welding. J Phys D: Appl Phys 2010; 43: 434001.

23.

Harmon

, et al. Laser-induced breakdown spectroscopy – an emerging chemical sensor technology for real-time field-portable, geochemical, mineralogical, and environmental applications. Appl Geochem 2006; 21: 730–747.

24.

Alfaro

SCA

Mendonça

DdS

Matos

. Emission spectrometry evaluation in arc welding monitoring system. J Mater Process Technol 2006; 179: 219–224.

25.

Hiraoka

Shiwaku

Ohji

. Determining temperature distributions of gas tungsten arc (TIG) plasma by spectroscopic methods. Weld Int 1997; 11: 688–696.

26.

Zhou

Heberlein

. Characterization of the arc cathode attachment by emission spectroscopy and comparison to theoretical predictions. Plasma Chem Plasma Process 1995; 16: S229–S244.

27.

Haidar

Farmer

AJD

. Temperature measurements for high-current free-burning arcs in nitrogen. J Phys D: Appl Phys 1993; 26: 1224.

28.

Olsen

. The electric arc as a light source for quantitative spectroscopy. J Quant Spectrosc Radiat Transfer 1963; 3: 305–333.

29.

Sousa Martins

, et al. Characterization of a high current pulsed arc using optical emission spectroscopy. J Phys D: Appl Phys 2016; 49: 415205.

30.

Lazorenko

Shapovalov

Kolyada

. ANALYSIS OF SPECTRUM OF THE WELDING ARC LIGHT FOR MONITORING OF ARC WELDING (Review). 2011.

31.

Zhang

, et al. Spectral diagnosis of wire arc additive manufacturing of al alloys. Additive Manufacturing 2019; 30: 100869.

32.

Dunn

Allemand

Eagar

. Metal vapors in gas tungsten arcs: part i. Spectroscopy and monochromatic photography. Metall Trans A 1986; 17: 1851–1863.

33.

Ren

, et al. In situ monitoring of Laser additive manufacturing for Al7075 alloy using emission spectroscopy and plume imaging. IEEE Access 2021; 9: 61671–61679.

34.

Zhao

Debroy

. Weld metal composition change during conduction mode laser welding of aluminum alloy 5182. Metallurgical and Materials Transactions B 2001; 32: 163–172.

35.

Manfredi

, et al. Additive Manufacturing of Al Alloys and Aluminium Matrix Composites (AMCs). In: Monteiro

(ed.) in Light Metal Alloys Applications. London, UK: IntechOpen, 2014.

36.

Zhao

White

DebRoy

. Current issues and problems in laser welding of automotive aluminium alloys. Int Mater Rev 1999; 44: 238–266.

37.

Cieslak

Fuerschbach

. On the weldability, composition, and hardness of pulsed and continuous nd:YAG laser welds in aluminum alloys 6061,5456, and 5086. Metall Trans B 1988; 19: 319–329.

38.

Hatch

Association

Metals

. Aluminum: Properties and physical metallurgy. USA: American Society for Metals, 1984.

39.

Song

, et al. Real-Time composition monitoring using support vector regression of Laser-induced plasma for Laser additive manufacturing. IEEE Trans Ind Electron 2017; 64: 633–642.

40.

Kramida

Ralchenko

Reader

, et al. NIST Atomic Spectra Database (version 5.11). 2023.

41.

Jagdish

Singh

SNT

. Laser-induced breakdown spectroscopy. 2nd Edition ed. Amsterdam, The Netherlands: Elsevier Science, 2020.

42.

Cremers

Knight

. Laser-induced breakdown spectroscopy. Hoboken, NJ, USA: Wiley Online Library, 2006.

43.

Manabu

Masao

. Plasma state in free-burning argon arc and its effect on anode heat transfer. J Phys D: Appl Phys 1999; 32: 1153.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.48 MB

0.00 MB