Computational Modeling-based Thermal Management for Pulsed Laser Welding of Small-Scale Devices

Pulsed laser welding is a widely used process to produce a hermetic seal for the enclosure of sensitive components in containers of small-scale devices such as those found in the medical and electronic industries. Through the nature of the welding process there is a local increase in temperature that must be dissipated through conduction of heat into the body of the component and convection and radiation into the surrounding environment. The resulting temperature increases may cause damage to some sensitive components.

Figure 1 shows a simple rectangular laser pulse waveform, where P_max is the peak laser power, t_on is the laser-on time, and t_a is the total time of a pulse. The pulsed laser welding procedure can be characterized by three parameters: pulse energy (P_max × t_on), pulse frequency (1/t_a), and travel speed. EWI has developed a computational modeling-based framework to accurately predict the temperature to which sensitive components may be exposed during welding.

Approach
The EWI thermal management framework is built upon advanced finite element heat flow analysis. The input to the model includes the geometry of device assembly, thermal-physical-mechanical property of materials, and laser welding parameters. The model takes into account the individual laser pulses, heat conduction into the body, heat convection and radiation into the surrounding environment, and thermal-mechanical contacts between different parts and the weld fixture. The evolution of temperature distribution in the entire device is predicted using the model.

Case Study - Pulsed Laser Welding of Flat Titanium Sheets
The usefulness of the thermal management model is demonstrated in the following case study. Two different sets of laser welding parameters are used to join 0.3-mm-thick flat, commercially pure, titanium plates. As shown in Figure 2, parameter set A has a slower travel speed and a lower pulse frequency than parameter set B. Figure 2 shows the distribution of peak temperature at 1 second after the start of welding. The weld pool (plotted in gray) for set A has a higher temperature than that for set B. This is expected since a laser pulse in set A has a higher peak power than that in set B.

Interestingly, it is noted that the heat-affected zone for set A is much smaller than that for set B. In other words, if a sensitive component is placed at the same distance from the weld center line, it will experience much lower temperature for set A than set B. This phenomenon is more clearly illustrated in Figure 3, where the thermal cycles at 0.6 and 1.0 mm away from the weld center line are plotted. As shown in this figure, the individual laser pulses lead to oscillation in the temperature profile. When the laser pulse is on, the temperature rises and vice versa. The closer to the weld center, the larger the temperature oscillation occurs. Since set A has a much lower pulse frequency, it provides sufficient time for the heat to dissipate into the plate. As a result, there is a lower heat build-up near the weld. On the other hand, set B constantly puts heat into the weld, and there is not enough time for the heat conduction away from the weld region. Therefore, there is extensive heat accumulation near the weld. The predicted thermal history indicates that set A can effectively decrease the peak temperature experienced by sensitive components located in the vicinity of the weld.

Summary and Conclusions
EWI’s heat flow analysis model can accurately predict the temperature profiles in simple geometries or in an entire device during pulsed laser welding. This allows for the improvement of welding procedurevelopment to reduce the risk of sensitive components which may be exposed to elevated temperatures.

For more information, please contact Wei Zhang at 614.688.5163 or wei_zhang@ewi.org.