Pre-welding analysis and associated laser welding methods and fiber lasers utilizing pre-selected spectral bandwidths that avoid the spectrum of an electronic transition of a metal/alloy vapor

Abstract

The present invention benefits from the determination that pre-selected spectral bandwidths that avoid the spectrum of an electronic transition of a metal/alloy vapor allow for welds substantially free from detritus that may discolor the weld. Accordingly, the present invention provides analytical methods, welding methods and fiber lasers configured to provide high quality metal/alloy welds.

Claims

1. A method of metal/alloy welding independent of the spectral properties of a cover gas, the method comprising: a. determining the spectrum of an electronic transition of a metal/alloy vapor associated with a workpiece; b. selecting a laser wavelength range and a linewidth range outside the spectrum of any electronic transition of the metal/alloy vapor, wherein the selected laser wavelength range and linewidth range does not interfere with any spectral lines of the metal/alloy vapor having an oscillator strength value above the standard deviation of the range of oscillator strength values for spectral lines of the metal/alloy vapor within a selected wavelength range; and c. subjecting the workpiece to laser output in the selected laser wavelength range and linewidth range to create a weld.

2. The method of claim 1 wherein the metal/alloy is titanium or a titanium alloy.

3. A method of creating a metal/alloy weld in a workpiece independent of the spectrum of a cover gas, the method comprising: a. selecting one laser wavelength range and linewidth range of a number of spectral bandwidths that avoid the spectrum of an electronic transition of a metal/alloy vapor associated with the workpiece, wherein the selected laser wavelength range and linewidth range does not interfere with any spectral lines of the metal/alloy vapor having an oscillator strength value above the standard deviation of a range of oscillator strength values for spectral lines of the metal/alloy vapor within the selected wavelength range and linewidth range; b. providing a spectrally-stable laser with power sufficient to create a weld pool with one of the spectral bandwidths; and c. subjecting the workpiece with an output from said laser sufficient to create a weld.

4. The method of claim 3, wherein the metal/alloy is titanium or a titanium alloy.

5. The method of claim 3, wherein subjecting the work piece with the output from the laser reduces the creation of coloration from plasma detritus.

6. The method of claim 3, wherein real time detected signals from the welding process are characterized by reduced signals coming from metal vapor plume laser light interaction.

Description

DETAILED DESCRIPTION

(1) FIGS. 1 to 8 illustrate different aspects related to the present invention, namely:

(2) FIG. 1 shows a weld-spot using a Nd:YAG (left) and off the shelf fiber laser 1070+−5 nm (right) in Argon atmosphere (1 kW peak power, 1.5 ms pulse duration)

(3) FIG. 2 shows a weld-spot using the modified narrow linewidth fiber laser (1064±0.5 nm)

(4) FIG. 3 is a high-speed image of a typical plume observed with the narrow band 1064±0.5 nm fiber laser. The dynamics within the plume are indicated by the arrows in the right picture.

(5) FIG. 4 illustrates a plume expansion of the standard 1070±5 nm fiber laser (left) and the narrow band 1064±0.5 nm fiber laser (right)

(6) FIG. 5 shows that no spot coloration was observed using a 1071.1±0.5 nm laser, hitting electronic transitions of Ti II (wherein Ti II stands for ions and Ti I stands for Ti atoms, as explained below) at 1070.7 nm as well as Argon at 1071.3 and 1071.5 nm (1 kW peak power, 1.5 ms pulse duration).

(7) FIG. 6 illustrates a plume shape and spot coloration observed using Zirconium and the standard 1070±5 nm fiber laser hitting eleven electronic transitions of Zr I (i.e. Zirconium atoms in the gas phase).

(8) FIG. 7 illustrates a plume shape and spot coloration observed using Zirconium and the narrow band 1064±0.5 nm fiber laser hitting no electronic transitions of Zr I.

(9) FIG. 8 illustrates a plume expansion and weld coloration using the standard 1070±5 nm fiber laser in Argon (left) and Helium atmosphere (right). The improved visual aspect is due to the unhindered plume expansion in case of the Helium atmosphere (lower density).

(10) FIG. 9 illustrates location of electronic transitions (open circles) of Ti I vapor listed in the NIST database (top) and the Kurucz linelist (bottom).

(11) FIG. 10 illustrates different laser emission profiles and location of Ti I electronic transitions (circles) in the 1060 to 1080 nm range with the respective oscillator strength (gf value) obtained from the Kurucz linelist.

(12) FIG. 11 illustrates different laser emission profiles and location of Ti I electronic transitions (circles) having an oscillator strength (gf value) above the standard deviation within the 1060-1080 nm range.

(13) FIG. 12 shows an optical schematic of the disclosed MM fiber oscillator and FIG. 12A shows an exemplary multimode active fiber which may be used therein.

(14) Applicants maintain that the research associated with the present invention demonstrates that the coloration effect on the prior art welds identified in the '626 app actually are a result from the dynamics of the plume expansion.

(15) The plume expands from the weld forming a vortex and the particles that cannot follow the vortex are deposited around the weld (as shown in FIG. 3).

(16) The direct comparison of the plume expansion between the laser giving the bad and good visual weld quality (standard 1070±5 nm vs. narrow linewidth laser 1064±0.5 nm laser) shows that the plume of the standard laser is much brighter and compact, forming one big vortex circulating above the weld material (FIG. 4). This results in increased particle deposition around the weld, leading to the poor visual quality.

(17) Spectral analysis of the emitted light of the laser plume have shown that no emissions from the cover gas (Argon) are observed. In addition, in the plume of the 1064±0.5 nm laser only species of neutral Titanium are present (Titanium atoms, Ti I). So the plume consists of Titanium vapor only (“vapor plume”).

(18) In the plume of the 1070±5 nm laser species of neutral Titanium (Ti I) as well as Titanium ions (Ti II) are present (“plasma plume”).

(19) The present invention provides that ionization of titanium takes place when using the standard 1070±5 nm laser and that the resulting plasma plume shows significantly different expansion properties compared to the vapor plume, resulting in increased particle deposition and thus weld coloration.

(20) To avoid unfavorable plume expansion dynamics, ionization of the plume has to be avoided.

(21) Ionization can take place by interaction of the laser light with electronic transitions of the vapor (in this case Titanium). Therefore, the energy levels of the electronic transitions have to meet the laser wavelength.

(22) The NIST atomic spectra database (cf. Kramida, A., Ralchenko, Yu., Reader, J., and NIST ASD Team (2015). NIST Atomic Spectra Database (ver. 5.3), [Online]. Available: http://physics.nist.gov/asd [2016, Aug. 18]. National Institute of Standards and Technology, Gaithersburg, Md.) lists 12 spectral lines for Titanium vapor (Ti I) and 7 spectral lines for Titanium ions (Ti II) falling into wavelength range of the off-the-shelf fiber laser (1065-1075 nm).

(23) This NIST atomic spectra database lists no spectral lines for Titanium vapor (Ti I) or Titanium ions (Ti II) falling into wavelength range of the narrow band laser (1063.5-1064.5 nm). Another example of the literature providing information on spectral lines relevant substances is a so-called Kurucz linelist which is available under http://kurez.harvard.edu/linelists/gfnew/. This list reveals additional information on the electronic levels of Titanium. While NIST database lists 12 spectral lines for Titanium vapor (Ti I) in the wavelength range of the off-the-shelf fiber laser (1065-1075 nm), the Kurucz linelist provides data on 99 lines in this range.

(24) Argon (cover gas) plays no role in the ionization of the plume (no Argon emissions observed). This is further supported by Outred (cf. M. Outred, Tables of Atomic Spectral Lines for the 10000 A to 40000 A Region, J. Phys. Chem. Ref. Data 7, 1 (1978)) who reports the presence of Argon lines at 1063.8 and 1064.0 nm which falls in the range of the narrow band fiber laser giving a good visual weld quality.

(25) To prove that the interaction of neutral Titanium (Titanium vapor) is responsible for the ionization and that interactions with Titanium ions (plasma) as well as the lines of the cover gas play a secondary role, an additional fiber laser with narrow emission between 1070.6-1071.6 nm was built by the Applicant IPG Photonics Corp.

(26) The NIST atomic spectra database lists no spectral lines for Titanium vapor (Ti I) and one spectral line for Titanium ions (Ti II) in this wavelength range.

(27) The above mentioned publication by Outred lists two spectral lines for Argon in this wavelength range. Thus, the laser hits all transitions considered relevant (cover gas lines, Titanium ions) in the '626 app, but a good visual weld quality could be obtained (FIG. 5).

(28) The generality of the principle (avoiding ionization because of laser vapor interaction) was strengthened by observing the same effect using a different material than Titanium. An example for such a material Zirconium (Zr).

(29) Outred (cf above reference) reports eleven spectral lines for Zr I falling into the wavelength range of the off-the-shelf fiber laser (1065-1075 nm). No lines are reported within the wavelength range of the narrow band laser (1063.5-1064.5 nm).

(30) FIG. 6 shows the plume shape and spot coloration using the off-the-shelf laser while FIG. 7 shows the plume shape and spot coloration of the narrow band laser (1063.5-1064.5 nm).

(31) The findings are in agreement with the previously shown results for Titanium.

(32) The above experiments allow the Applicants to draw a number of conclusions.

(33) The coloration originates from the deposition of Titanium particles around the weld.

(34) Particle deposition depends on the plume expansion: A fast plume expansion gives a better visual quality than a slow plume expansion.

(35) The interactions of the laser emission with the electronic levels of Titanium vapor (neutral atoms, Ti I) leads to ionization of Titanium, forming Titanium ions and free electrons. This plume can now be called plasma.

(36) In comparison to the vapor plume, the plasma plume shows a different expansion behavior: it is more compact and expands very slowly. As a result, more particles are deposited around the weld, causing the strong coloration.

(37) Thus, ionization (creation of a plasma plume) of the plume has to be avoided.

(38) Therefore, the spectrum of the laser output should not interfere with the spectrum of any of the electronic transitions of the vapor (neutral atoms) of the material being welded. When the laser wavelength and energy levels do not match, no interaction and thus no photoionization will take place.

(39) However, review of an alternative database than the NIST Atomic Spectra Database, for example the Kurucz linelist, reveals additional information on the electronic levels of Titanium. While NIST lists 12 spectral lines for Titanium vapor (Ti I) in the wavelength range of the off-the-shelf fiber laser (1065-1075 nm), the Kurucz linelist provides data on 99 lines in this range. As a consequence, it would be extremely difficult to provide a fiber laser which does not hit any of those transitions (as shown in FIG. 9). In such situations, a so-called absorption efficiency, which is also known as oscillator strength, becomes important:

(40) It is possible that some electronic transitions absorb the laser radiation more efficient than others. So it could happen that a good visual weld quality is obtained despite electronic transitions being hit. However assessing this absorption efficiency requires extensive investigations. So avoiding every transition (if possible) is just the easiest way. Since the oscillator strength (log gf) is also reported in the Kurucz list, the identification of important lines becomes possible.

(41) For example, a fiber laser may be tuned in the 1060-1080 nm range for Ti welding. With all the data known from literature, there is no “gap” of electronic transition where the fiber laser could be tuned to (FIG. 9). Thus, important transitions have to be identified by laboratory experiments or by using data available in literature (determination of oscillator strength). In case of Titanium, oscillator strength values can be obtained for the Kurucz list (FIG. 10). An important line may now be defined as a line with a gf-value (linear scale, not log gf) above the standard deviation of all gf-values within the selected wavelength range (FIG. 11). Finally, the emission profile of the laser is modified to avoid those “important” wave-lengths as much as possible.

(42) Therefore, if selecting a laser wavelength and linewidth outside the spectrum of any electronic transition of the metal/alloy vapor is not possible, laser wavelength and linewidth is modified so according to this invention that way that interactions with lines of high oscillator strength are minimized.

(43) Interactions of the laser with the ions play a secondary role, because the ions have to be present to make this interaction happen. However, when they are present, it's already too late.

(44) No evidence for an interaction between the laser/cover gas could be found.

(45) Nevertheless the cover gas will influence the plume dynamics because of differences in gas density. For example Helium (10× lower density Argon) allows the plume to expand more easily and gives a good visual weld quality (see FIG. 8). But this has nothing to do with a laser/gas interaction.

(46) Finally FIG. 12 illustrates a multimode (“MM”) fiber laser or oscillator 10 configured with a gain medium which includes a MM fiber 12 doped with light emitters which can be used in accordance with the present invention. As known, the light emitters include ions of rare earth elements selected from ytterbium (“Yb”), erbium (“Er”), neodymium (“Nd”), thulium (“Tm”), holmium (“Ho”), praseodymium (“Pr”), cerium (“Ce”) yttrium (Y3+), samarium (Sm3+), europium (Eu3+), gadolinium (Gd3+), terbium (Tb3+), dysprosium (Dy3+), and lutetium (Lu3+) and various combinations of these.

(47) The active step index fiber 12 is configured with a core capable of supporting a great number of transverse modes typically associated with a core diameter greater than 20. As one of ordinary skill in the fiber laser art knows, a step-index fiber cannot continue to support a single mode once its core diameter exceeds a 30 μm diameter. The configuration of active fiber 12 further may have one or more claddings which surround the core in a manner well known to one of ordinary skill.

(48) The oscillator 10 is further configured with two MM passive fibers 14 fused to respective opposite ends of active fiber 12. The MM passive fibers each are configured with a core diameter and numerical aperture substantially matching those of the active fiber 12. A pump 20 includes one or a plurality of MM pigtailed laser diodes arranged in a side-pumping scheme and having an emission peak which corresponds to the absorption peak of the selected doping ions. The combination of passive, active and pump fibers combined as shown in FIG. 12 and, optionally, placed in a housing (not shown) constitutes a single gain block.

(49) The laser 10 has a Fabry-Perot configuration with a resonant cavity defined between spaced strong and weak reflectivity MM fiber Bragg gratings (“FBGs”) 16 and 18, respectively. The FBGs 16 and 18 are written in respective MM passive fibers 14. While FBGs can be written in the active fiber, it may be technologically challenging.

(50) Multi-kilowatt implementation of single mode (“SM”) high power fiber lasers has been achieved and is already available on the market. The state of the art demonstration of continuous wave SM operation is currently about 10 kW and prediction of SM operation levels exceeding 30 kW have been made. While the capability of fiber lasers to maintain an excellent beam quality at high power is undisputed, there is little concern of the laser linewidth as they generally used for processing of materials that have broad absorption band.

(51) However, certain applications require a selective spectral line. The spectral linewidth of the laser light should to be narrowed to avoid the absorption peak which may be realized e.g. by a high cost, single frequency SM fiber laser.

(52) As known, one of the main reasons for line broadening in SM fiber lasers is the presence of nonlinear effects (“NLE”), such as Raman scattering and four wave mixing, which become more pronounced with the power increase. The threshold for NLE can be lowered by increasing a core diameter and cavity length. Hence MM fibers characterized by large diameter cores and smaller cavity lengths are characterized by relatively narrow spectral linewidth.

(53) The active MM fiber 12 of the present disclosure fully meets the requirement for a relatively high NLE threshold. In one embodiment, the configuration of the MM active fiber 12 may include a typical cylindrical fiber. In the other embodiment, the active fiber may be configured with a double bottleneck shape, as shown in FIG. 12A, which has two relatively small diameter input and output core regions 11, a central relatively large diameter central amplifying core region 13 and two tapered regions 15 bridging the opposite ends of the central region with respective opposing ends of the ends core regions. The increased diameter of amplifying region is instrumental in a further increase of core diameter, reduction of fiber length and ultimately elevation of NLE threshold and narrowing of spectral linewidth. The cladding of MM active fiber shown in FIG. 12A may be configured with a typical cylindrically shaped cross-section or may have the same double bottleneck shape as the core. The spectral linewidth that would meet many industrial requirements, such certain types of material laser processing, for the disclosed laser may vary from 0.02 nm to about 10 nm. The desired spectral line may be obtained by careful selection of, among others, the core diameter, length of resonant cavity, configuration of MM Bragg gratings, dopant concentration.