A high power direct diode [HPDDL] laser and its unique beam make for a highly efficient tool to use in cladding operations. Laser cladding is performed by melting a pre-placed powder onto a substrate to ensure a bond with minimal dilution, nominal melting and a small heat affected zone. The laser used in the experiment was a 4000 watt Direct Diode laser system (HPDDL) mounted on a Panasonic VR-16 robot. The pre-placed powders chosen for this experiment are ANVAL 410, 156 and C22. 410 and C22 were selected for their superior corrosion resistance. 156 is a general-purpose cobalt hard facing material. The cladding substrate was ASTM 1018 steel. The dilution of the coatings was analyzed through the use of a Scanning Electron Microscope [SEM]. Through analysis it was discovered that dilution is kept to a minimum, in the range of 0 to .02%. The corrosion resistance and wear resistance was also measured for the appropriate samples. This process is highly advantageous in comparison with competing coating methods such as plasma spraying, arc welding, and other laser sources. The rewards being lower porosity reduced post-machining, optimum edge detail.
As tools for use in industrial applications, HPDDL, also known as semiconductor lasers, are becoming more prevalent.1,2,3 Diode laser technology has been used for a number of years in compact disks, laser printers and laser pointers. Their low cost, high efficiency, and compact design make them an attractive technology in the industrial manufacturing environment. The electrical to optical conversion efficiency of the HPDDL is as high as 55%.
The light emitted at the facet of the laser diode is highly divergent and astigmatic. To make this usable, a lenslet array is close coupled to a two dimensional array of laser diodes. Since the other axis, referred to as the “slow axis,” is not collimated and is left to diverge, the final focusing lens will produce a concentrated line of light, which is very useful for large area applications such as cladding. This beam is very uniform, having a nearly tophat intensity profile along the long axis with a guassian profile perpendicular to the line along the short axis. The HPDDL used in this feasibility study employs 4 stacks of 20 bars, which are brought to a line by a single macro lens [Figure 1]. With dimensions of approximately 12.5 mm X < 1 mm with a 125 mm focal length lens. With different macro lenses this laser can achieve power densities greater than 200 kW/cm2.
Figure 1 Focus Configuration of Line Source HPLLD
An ideal application for the HPDDL is large surface area laser cladding. As shown in Figure 1 the line of laser light along the short axis is moved perpendicular to the long axis. The biggest benefit of HPDDL laser cladding is that the unique line source allows the user to produce clads with a controllable width without scanning. CO2 and Nd: YAG lasers have a smaller spot; thus the laser must be scanned over the cladded area. The wavelength of the HPDDL is 808 nm, compared with 1.06 _ m for a Nd: YAG laser and 10.7 _ m of the CO2 laser. The shorter wavelength of the HPDDL allows for higher absorption into the material being cladded, therefore a higher process speed can be achieved. Both CO2 and Nd: YAG lasers often require binders when using pre-placed powders. The use of binders often leads to porosity due to the evaporation of volatiles during the cladding pass4. The HPDDL system does not necessitate the use of binders to hold the powder together before a cladding pass. Another advantage of the HPDDL is that the thermal input can be precisely controlled thus yielding minimal dilution and a small heat affected zone.
During a laser cladding process dilution is expected to be minimized. In cladding operations dilution is often defined as the amount of intermixing of the clad and substrate. Dilution is measured by visual analysis or through a SEM elemental line scan. Visual analysis allows the user to get a quick estimate of the dilution of the clad, however this method of measurement is not very accurate. Through visual analysis dilution is defined as the distance the clad layer extends below the substrate. SEM analysis is a true, accurate measure of the dilution, or intermixing of the clad and substrate. Laser alloying is a process that is often grouped with laser cladding operations. Laser cladding and alloying are traditionally distinguished by the relative amounts of the consumable material added and substrate melted. Generally the two categories are arbitrarily separated by their relative amount of dilution, laser alloying being classified as having greater than 10% dilution, laser cladding having less than 10% dilution4. In laser alloying it is generally desired to mix portions of the coating with the substrate to produce an alloyed layer, thus a high dilution and high intermixing is expected. It should also be noted that laser alloying requires convection and laser cladding does not. In many laser alloying processes the cooling rate is often monitored to ensure intermixing and the formation of unique metallurgical compounds. Ultrafast quench rates of the order of 1011 Ks-1 are often required4 as well as a high solubility of the clad material in the parent material. Laser alloying experiments were not conducted in this study, however throughout the experimentation there was an expectation that at a low process speed some alloying of the powder and substrate may occur. This was not true for the HPDDL process because laser alloying requires very high quench rates and a keyhole as seen in Nd:YAG and CO2 lasers.
The denser microstructure and better bonding of laser clads allows for enhanced corrosion and wear resistance with a single pass. Laser cladding is a viable alternative to plasma spraying and TIG or MIG processes. The clad material deposit does not intermix with the substrate in many applications; therefore the dense, uniform microstructure of the clad layer allows for enhanced single pass corrosion or wear resistance in a HPDDL clad. It is difficult to produce a clad with a TIG, MIG or plasma spray system without having less than 5% dilution, therefore as many as 15 overlapping passes are required to obtain an undiluted clad layer5. Conventional arc welding processes generally impart a significant amount of heat into the part resulting in a large heat affected zone and distortion. Post-weld treatment can improve the properties of the joint, but can also lead to distortion of the component6. The surface finish of overlapping passes produced with the HPDDL are relatively flat, however a TIG cladding process often results in distinct ridges and valleys, which lead to cracking when bent7. In addition, the arc welding processes often are also responsible for the losses of alloying elements8. A direct comparison of a laser clad layer with an arc-welded layer indicates that the HPDDL clad has significant grain refinement, which in some cases lead to an increased wear resistance9. The HPDDL also surpasses flame spray technology, since flame spray produces a more porous coating with limited adhesion10.
Laser cladding also has several advantages over plasma cladding processes. The substrate of laser clads are free of the micro-cracks and pores typical with the plasma clad process. Other advantages of the HPDDL over plasma processes include the uniformity of the HPDDL coating, the manual requirement of plasma processes, cracks and pores in a plasma clad. The sharp boundary of the plasma clad layer with the substrate also often leads to pores and cracking9. The interface between the clad and substrate of a HPDDL clad is smooth with minimal dilution.
Multiple pass samples were prepared which demonstrated uniform cladding thickness [Figure 3]. Recent research has been performed on 100% overlapping clad passes that indicate that this significantly increase the cladded surface properties10. Corrosion testing indicated that the overlapping passes could withstand prolonged salt spray exposure. Surface roughness and uniformity of the clad are two important properties that are influenced by overlapping clads4. Overlapping passes result in a decrease in surface roughness and are typically dense and well bonded.
The properties of the clad material alone will not determine the properties of the clad on the substrate. The solubility of the clad, which determines the amount of intermixing of the clad and substrate, i.e. dilution, is important. The resulting microstructure of the clad, the dilution layer and heat-affected zone are all important areas in determining the quality of the clad. Finally, solubility and wetting issues that lead to pits and pores. All of the above influence the wear and corrosion resistance of the clad.
Wear and corrosion resistant powders were selected for the experimentation. The corrosion resistant powders include C22 is a NiCrMo alloy in the Hastealloy C family, and 410 is a basic stainless T410 material. The nominal composition of each alloy is listed in Table 2. The substrate used was 1018 steel, which was selected because it is a commonly used and inexpensive material.
The 156 material is a cobalt based hardfacing alloy used for increased wear resistance. The composition of this alloy consists mainly of cobalt, however Cr is also largely alloyed in this material [Table 2].
Material / Performance Evaluation
The powder was pre-placed to a thickness of .050″ on a 1018 steel substrate. The thickness and width of the cladding pass changes with modifications in processing speed. As the processing speed increases the clad track has an increasingly gaussian profile due to the surface tension of the melt. However, a decrease in speed results in a flatter, wider clad with high visual dilution [Figure 2]. Overlapping passes wet together to form a relatively flat profile regardless of processing speed .
Figure 2: A comparison of the profiles of two NiCrMo clads. The clad on the right was produced at a travel speed of 0.45 m/min, the clad to the left at a process speed of 0.70 m/min.
Experimental Procedure The HPDDL was used to clad the pre-placed powders onto the substrate. The line source was passed along the short axis over the powder. The speeds varied from 0.3 to 0.8 m/min at 4 kW of laser power. The variance in the speed allowed for clads to be produced with varying levels of visual dilution. Each powder was cladded with a visual dilution of 0, 10 and 60%. Two clads were produced for each dilution level.
One of the two clads was sent for SEM analysis, one of the hardfacing clads at each dilution level was sent for wear testing and the corrosion resistant clads were used for corrosion analysis. SEM analysis was completed on all of the samples to determine the level of dilution and change in dilution with overlapping passes. The profile of these samples was a relatively flat surface. Corrosion testing was done on the stainless steel samples. This test was performed by immersing the samples in nitric acid for a period of 24 hours to determine the effect of the acid on the substrate and clad. Corrosion testing was also completed on the NiCrMo alloy by immersing the clad and substrate in a phosphoric acid solution. Wear testing was done on the Cobalt based clad layer. The standard pin on disk test was done in accordance with ASTM G99 to determine the resistance to galling of the clad. A water-jet test in accordance with ASTM D5367-94 was also performed to determine the wear resistance of the clad. Multiple pass samples were also produced.
126.96.36.199 Visual Examination Visual measurement of dilution was performed by using the substrate as a base for all measurements. As the clads were produced with the HPDDL a portion of each clad was cut off, ground with 180 grit paper, and etched in 2% Nital to determine the visual dilution. The portion of the clad that was above the substrate was measured at the highest point as well as the entire length of the clad layer. The portion of the clad below the substrate was divided by the length of the total clad layer to produce a percentage visual dilution [Figure 4]. The initial dilution measurements described above are shown in Table 2. The drawback to this method of measuring dilution is the lack of accuracy in measurements. However, visual dilution measurements are a straightforward approach to determining the approximate dilution of a sample while processing.
Figure 4: Visual measurement of dilution was performed through the equation L2/L1.
Table 2: Dilution as measured by visual inspection.
An acid etch was performed on each of the samples to bring out the microstructure of the clad layer. The etch used for the 410T stainless material was oxalic acid, while the NiCrMo and cobalt based alloys were etched electrolytically in a solution containing equal amounts of CrO3 and potassium permanganate, and 8% sodium hydroxide. The microstructures indicate thorough melting of the powder. Both the NiCrMo and Cobalt based alloy show a dendritic microstructure within the clad layer [Figure 5]. Grain growth is seen in the heat affected zone of the clad, however there is no evidence of the melting of the substrate. The 410 T SS powder shows also shows grain growth in the heat affected zone, but the microstructure of the clad show is primarily martensitic due to the rapid quench rate of the powder [Figure 6]. The microstructures present indicate that the dilution of the clad into the substrate is minimal and that changes in process speed do not reflect changes in dilution.
Figure 5: Dendritic formation in the Cobalt based clad layer , also the interface between the clad and substrate is shown on the left.
Figure 6: Martensitic formation in the 410T Stainless clad layer , also the interface between the clad and substrate, left.
SEM Analysis A SEM line trace was used on each of the samples to determine the dilution of the clad layer as defined by the amount of intermixing of the clad layer and substrate. Each powder has a reasonable amount of Chromium; therefore this element was traced in the clad layer for each powder. Iron was traced in the substrate.
At a high process speed the dilution of the clad into the substrate is minimized. The cobalt based hardfacing powder was clad at a speed of 0.7 m/min at a power of 4 kW
At a lower processing speed the dilution is still minimized. A clad was produced at a speed of 0.40 m/min at 4 kW with the cobalt based hardfacing powder
The properties of overlapping passes with regard to dilution and amount of intermixing are similar to those of a single pass. The same cobalt based powder has minimal dilution and intermixing at a process speed of 0.7 m/min, 4 kW
Samples produced with the stainless steel and NiCrMo powders produced similar results with respect to dilution. At all of the process speeds the dilution was minimal. As overlapping passes are produced to create a 100% clad surface, no effect on dilution is observed.
Corrosion Testing Corrosion testing was performed on the stainless steel samples by immersing the clad and substrate in nitric acid for a period of twenty-four hours to determine the effect of the acid on the substrate and clad. Corrosion testing was also completed on the NiCrMo alloy by immersing the clad and substrate in a pure phosphoric acid solution for twenty-four hours.
The corrosion analysis indicates that a great deal of corrosion occurs for each of the 410 Stainless Steel samples. However, this analysis is also somewhat misleading. The entire clad and substrate was immersed in the acidic solution, the majority of the corrosion occurred in the substrate. In most industrial applications only the clad would be exposed to the corrosive media. A visual of analysis of the clad and substrate of the single pass stainless steel samples before and after the corrosion testing indicates that the clad is virtually unaffected, however the substrate has dissolved in the acid [Figure 11]. Overlapping passes produced similar results with the majority of the material loss being in the substrate [Figure 12]. The 410 Stainless Steel clads also have a visible change in color after immersion in the nitric acid. This is an indication that the passive Cr2O3 layer has been removed thereby increasing the corrosion rate.
Figure 11: A visual comparison of the 410 Stainless Steel clad produced at 4kW, 0.65 m/min before and after immersion in nitric acid indicates that most of the material loss is in the substrate
Figure 12: Overlapping passes produced at 0.65 m/min, 4 kW show the majority of material loss being in the substrate.
The C22 alloy was fairly resistant to the phosphoric acid. The clad layer is unaffected in this acidic solution, however pitting can be observed in the substrate. A change in the color of the substrate is also observed, this indicates the beginning of a loss of the normally present thin film of iron oxide in the steel.
Pin on Disk To produce consistent values for relative wear resistance, a standard pin-on-disk wear testing machine was used in accordance to ASTM standard G99. The data indicates that with a slower processing speed the wear resistance will slightly increase, or the percent mass loss will decrease [Figure 13]. The 0.40 m/min observed a slightly lower mass loss than that of the samples produced at faster speeds. The decrease in mass loss with a decrease in speed is due, in part, to the denser microstructure produced at a slower speed. The overlapping passes also have a slightly lower loss of material than the single pass samples. The decrease in mass loss is not significant enough to draw reasonable conclusions. However, this may be a slight indication that the overlapping passes have superior properties than single pass samples due to increased surface roughness and the denser microstructure. It is also visible from figure 13 that the mass loss of all of the clad layers is significantly less than that of the substrate.
Ablation Testing The further wear testing of the clad layer was performed using a water-jet and scanning the 100 grit garnet fluid over the top of the clad layer and substrate at a speed of 2.54 m/min [Figure 10]. The pressure of the water-jet was at 344 MPa, the stand off of the jet was 0.0127 m. The change in thickness from the original clad profile was measured and recorded. The percent material loss was determined by the equation:
(tcontrol sample tablated sample)/ tcontrol sample 
The results were recorded and a direct comparison can be made between the material loss in the clad layer and the material loss in the substrate [Figure 14]. It was found that the 1018 steel generally experienced a greater loss of material than the Cobalt based alloy. However, the 0.75 m/min clad was subjected to a higher degree of mass loss than the base material.
The profile of the clad and substrate was examined and compared to a control specimen from the same cladding pass that received no treatment. Visual examination indicates that there is a substantially greater loss of material in the substrate than in the clad layer
Figure 14: Measured loss of material in the substrate and clad. The bottom picture shows a comparison of the profiles of the Cobalt based clad layer produced at 0.7 m/min at 4 kW. The picture to the left is untreated, the clad on the right has been water-jet wear tested.
Results and Discussion
Through analysis it was discovered that the clads produced with the HPDDL performed well throughout various tests. The SEM analysis indicated that regardless of process speed, the HPDDL clads had minimal dilution. It was also found that overlapping passes also had minimal levels of dilution. Metallographic analysis indicated that thorough melting of the clad layer occurred, as well as rapid quench rates were observed. The stainless steel and NiCrMo samples were sent for corrosion testing. The immersion of the stainless steel clad in nitric acid resulted in the dissolving of the substrate. The NiCrMo clads, when immersed in phosphoric acid, simply lost their passive layer and experienced some pitting. Pin on disk wear analysis showed that the wear due to galling was similar for each of the cladded samples. The abrasive wear analysis indicated that the clad layer is more resistant to abrasion than the substrate, as would be expected
Through experimentation it was found that the HPDDL is an effective method of producing high quality clads with minimal dilution. It was found that the corrosion and wear properties of HPDDL clads are equal to those produced with competing methods such as plasma spray, TIG or MIG deposits. The HPDDL allows the user to produce a single pass clad with minimal dilution. This can not be accomplished by traditional arc welding processes, which require multiple passes to achieve a pure clad layer. The low dilution clads with controllable thickness are beneficial because the end user can save the expensive of purchasing excessive amounts of expensive cladding wire and powder. Laser cladding is highly advantageous over TIG and MIG processes because the amount of dilution is controllable, it is an automated process, chemically clean and environmentally friendly. The primary advantage of the HPDDL in comparison to CO2 and Nd:YAG lasers is the shorter wavelength and thus higher absorption of the direct diode laser. Other benefits of the HPDDL over conventional lasers are the elimination of scanning, controllable dilution and the elimination of binders with pre-placed powders. The HPDDL is a highly capable cladding tool that will produce coatings with first-rate corrosion and wear resistance, low dilution, low porosity, unique microstructures and aesthetic surface finishes.
- P. Loosen et. al., SPIE, 2382, 78-87 (1987)
- S. Pflueger et. al. “Material Processing with high Power diode lasers” Automotive Laser Applications. 1995 workshop
- S. Pflueger and F. Kuepper, ESD Technology, April/May 1996
- S. V. Joshi and G. Sundararajan in N. Dahotre, ed. Lasers in Surface Engineering, ASM International, Ontario, 1998, pp.121-124, 139-144, 149-153.
- C. L. Horn et. al. in T. Lyman. Metals Handbook: Welding and Brazing, American Society for Metals, Metals Park, 1981, pp. 149-161.
- K. C. Meinert, Jr. and P. Bergan, ICALEO 1999 Proceedings, 87, F49 (1999).
- T. Heston, Welding Journal, 79 (7), 46 (2000).
- H. Ocken, Advanced Materials and Processes, 157 (6), 103 (2000).
- B. Medres, L. Shepeleva and M. Bamberger in ref. 6, pp. F225-F230.
- R. Hull et. al. in ref. 6, pp. 41,45-47.
- ASTM Standard G31-72 (1999) ASTM Subcommittee G01.05.
Crystal M. Cook, Applications Engineer. Ms. Cook graduated from the University of Missouri Rolla in 1999 and joined Nuvonyx, Inc. in January of 2000.
John M. Haake, Vice-President Market Development. Mr. Haake graduated from the University of Missouri Rolla in 1988 and worked at McDonnell Douglas until 1998. Mr. Haake has over 14 patents in relation to laser diode technology.
Surface Transformation Hardening of Gray Cast Iron
Surface transformation hardening of Class 40 Gray Cast Iron was also performed without the use of an absorptive coating or inert gas shielding using a HPDDL. The beam was defocused 4 to 8 mm and moved at various speeds over the work-piece. The gray cast iron was much more sensitive than the 4140 due to a very narrow process widow between melting, producing surface carbides, and desirable hardening. It was determined that at a 6 mm defocus a case would be produced with a higher hardness value than the case produced at 8 mm defocus. The 4 mm defocus produced surface melting and carbides. At a higher degree of defocusing the case depth would increase at the expense of hardness. This will occur up to a critical amount of defocusing. This application will benefit greatly from temperature control, which is completely within the means of the HPDDL.