Fig. 4 — Macroscopic cross-sections of laser welds made on Type 304L stainless steel pipe during radiation pressure experiments. Real time, RPM-measured laser welding power (kW) and linear heat input HI (kJ/mm) are shown for each weld. (a granite slab) of the welding workstation so it did not move with the translation stage on which the workpiece was mounted. The scale was protected by an aluminum housing (Fig. 2) with an anodized aluminum tube containing the focusing lens and a disposable cover glass to protect the lens from damage by weld splatter. Finally, the workpiece was placed at the focus of the laser light, ~300 mm from the lens. Nitrogen shielding gas was delivered from a nozzle placed approximately 2 cm from the workpiece. We found that heat from the weld pool and plume affected the scale’s measurement, so a foil heat shield was placed between the weld location and the scale housing with a hole for the light to pass through. Of course, this setup is not typical for laser welding operations, but it allows us to test the performance of our prototype RPM. Further developments will be needed to miniaturize it to reside in the laser weld head itself and to implement noise rejection techniques allowing operation in a high-vibration environment. Demonstration of Real-Time Radiation Pressure Technique To demonstrate real-time laserpower measurement during the welding process, we performed several circumferential welds on a Type 304L stainless steel pipe of 89.5 mm outer diameter and a wall thickness of 5.6 mm. Nitrogen shielding gas was used to provide an inert environment and the beam was focused to a spot size of ~0.6 mm diameter at the workpiece. Travel speed (pipe rotational speed) was increased with increasing laser power to maintain good weld quality based on visual surface inspection. After welding, the pipe was crosssectioned, polished, and etched with mixed acid (equal parts HCl, HNO3, and acetic acid) to reveal the macrostructure using optical microscopy. Figure 4 shows the cross-sections, which reveal weld penetration depth as a function of laser power as measured during the weld using the RPM. As expected, the weld penetration depth and total melt volume increased with laser power, reaching complete joint penetration at 2.8 kW beam power. Note that the linear heat input, HI, was calculated for each weld since pipe rotational speed ω (in units of mm/s) varied according to the selected laser beam power, P, so that HI = P/ω. The uncertainty of the laser power measurement using the radiation pressure power meter is a preliminary estimate and will be refined further with more measurements. But currently, the uncertainty is dominated by the uncertainty in our calibration of the scale over the lowest mass ranges (300–500 μg), which correspond to about 600–1000 W of laser power. We estimate that the uncertainty of the scale calibration at these lowest levels is about 1.5%. For now, we use this as our power uncertainty estimate. We are also aware of a potential thermal drift in the scale reading as higher powers or longer weld times change the temperature of the RPM. These effects must be addressed, but for the parameters measured here, any drift was simply removed by a linear approximation and we assigned a tentative laser power measurement uncertainty of 1.5%, which includes a coverage factor of 2 (sometimes referred to as a “2 sigma uncertainty”) indicating that we expect the actual power has a 95% probability that it is within 1.5% of the value measured by the RPM. Hybrid Power Measurement Technique to Achieve Accuracy and Speed As mentioned, radiation pressure is a unique way to measure laser power because all other methods require the laser light to be absorbed. Traditionally, the tradeoff is between absorbing all the light for an accurate power measurement, or alternatively, measuring only a tiny “pick-off” fraction of the light allowing the rest of the light to be used for the welding operation. This second approach has lower potential accuracy but is a simple technique offering a fast way to see changes in laser power during a weld with a response time on the order of milliseconds or even microseconds. Current commercial scale technology, on the other hand, is not designed for such rapid measurements and as a result, our RPM has an approximate 5-s response time. In the short term, a hybrid approach between a radiation pressure power meter and a pickoff power monitor might be a solution to enable both rapid and accurate real-time measurements of welding laser power. The pickoff approach is somewhat common and came installed in the fiber laser feeding our welding operation. In the measurement, a small fraction of the laser’s light is absorbed for measurement using a photodiode (a small, semiconductor-based optical power detector). This device works like a small solar cell where light is ab- MARCH 2016 / WELDING JOURNAL 33
Welding Journal | March 2016
To see the actual publication please follow the link above