دانلود رایگان مقاله انگلیسی مدل سازی اجزا محدود برای تغییر شکل فیوژن پودر لیزری تخت - نشریه الزویر

abstract

Experimental measurements are a critical component of model development, as they are needed to validate the accuracy of the model predictions. Currently, there is a deficiency in the availability of experimental data for laser powder bed fusion made parts. Here, two experimental builds of cylindrical geometry, one using a rotating scan pattern and the other using a constant scan pattern, are designed to provide post-build distortion measurements. Measurements are made using a coordinate-measuring machine which provides distortion profiles along the height of the part at four separate locations. Measurements show thatfor these cylindricalthin wall builds,there is no discernable effect on distortion from using the rotating versus constant scan patterns. Project Pan finite element modeling software is used to model each of the experimental builds. The simulation results show good agreement with experimental measurements of post-build deformation, within a 12% percent error as compared to experimental measurements. Using the FE model, the effect of a flexible versus a rigid substrate on distortion profile is examined. The FE model is validated against in situ experimental measurements of substrate distortion. The simulated results are used to study stress and distortion evolution during the build process. Internal stresses calculated by the model throughout the part are used in explaining the final part distortion. The combination of experimental and simulation results from this study show that the distortion of the top layer is relatively small (less than 30%) throughout the duration of the build process compared to the peak distortion, which occurs several layers below the most recently deposited layer. For these geometries once the part is built to a sufficient height, the peak distortion magnitude does not change.

5. Conclusion

Two experimental builds with a simple cylindrical geometry are manufactured in a LPBF machine, one built with a rotating scan pattern and the other with a constant scan pattern, with the goal of providing measurements of the post-build distortion. CMM measurements of distortion show that for both cases, the parts distort towards the center line of the cylinder with approximately equal magnitude for each measurement location for both cases. Experimental measurements are compared against Project Pan FE model for validation purposes. FE model results compare well with experimental measurements made. The highest averaged percent error for any measurement location distortion profile falls within 12% of the experimental measurement. Project Pan has been demonstrated effectively model in situ distortion ofthe substrate in the direction ofthe build height(Z) within 10%. By completing simulations using both a rigid and flexible substrate, it was determined that flexibility in the substrate affects the shape of the distortion profile for the X-direction. The inclusion of theflexible substrate for these simulations increases accuracy inthe X measurement locations by 2–3% albeit at the cost of accuracy in the Ydirections.Additionalutility ofthemodelis alsodemonstrated in extension of the rotating scan pattern case (Case 1), which was under-built due to complication during the build process. The FE model first validated on the experimental build of Case 1, extended the build height of that case to the predefined build height for comparison with the constant scan pattern (Case 2). Comparison of the model results for these cases show little discrepancy, further demonstrating that a rotating versus constant scan pattern does not have a significant impact on these geometries.