DETECTION OF DENSE, OVERLAPPING, GEOMETRIC OBJECTS.

Using a unique data collection, we are able to study the detection of dense geometric objects in image data where object density, clarity, and size vary. The data is a large set of black and white images of scatterplots, taken from journals reporting thermophysical property data of metal systems, whose plot points are represented primarily by circles, triangles, and squares. We built a highly accurate single class U-Net convolutional neural network model to identify 97 % of image objects in a defined set of test images, locating the centers of the objects to within a few pixels of the correct locations. We found an optimal way in which to mark our training data masks to achieve this level of accuracy. The optimal markings for object classification, however, required more information in the masks to identify particular types of geometries. We show a range of different patterns used to mark the training data masks, and how they help or hurt our dual goals of location and classification. Altering the annotations in the segmentation masks can increase both the accuracy of object classification and localization on the plots, more than other factors such as adding loss terms to the network calculations. However, localization of the plot points and classification of the geometric objects require different optimal training data.

Time-Resolved Absorptance and Melt Pool Dynamics during Intense Laser Irradiation of a Metal.

High-irradiance lasers incident on metal surfaces create a complex, dynamic process through which the metal can rapidly change from highly reflective to strongly absorbing. Absolute knowledge of this process underpins important industrial laser processes such as laser welding, cutting, and metal additive manufacturing. Determining the time-dependent absorptance of the laser light by a material is important, not only for gaining a fundamental understanding of the light-matter interaction but also for improving process design in manufacturing. Measurements of the dynamic optical absorptance are notoriously difficult due to the rapidly changing nature of the absorbing medium. These data are also of vital importance to process modelers, whose complex simulations need reliable, accurate input data; yet, there are very few available. In this work, we measure the time-dependent, reflected light during a 10-ms laser spot weld using an integrating-sphere apparatus. From this, we calculate the dynamic absorptance for 1070-nm-wavelength light incident on 316L stainless steel. The time resolution of our experiment (less than 1 µs) allows the determination of the precise conditions under which several important physical phenomena occur, such as melt and keyhole formation. The average absorptances determined optically are compared with calorimetrically determined values, and it is found that the calorimeter severely underestimates the absorbed energy due to mass lost during the spot weld. Weld-nugget cross sections are also presented to verify our interpretation of the optical results, as well as to provide experimental data for weld-model validation.

Experimental observation of extremely weak optical scattering from an interlocking carbon nanotube array.

We experimentally demonstrate a nearly wavelength-independent optical reflection from an extremely rough carbon nanotube sample. The sample is made of a vertically aligned nanotube array, is a super dark material, and exhibits a near-perfect blackbody emission at T=450 K-600 K. No other material exhibits such optical properties, i.e., ultralow reflectance accompanied by a lack of wavelength scaling behavior. This observation is a result of the lowest ever measured reflectance (R=0.0003) of the sample over a broad infrared wavelength of 3 µm < λ < 13 µm. This discovery may be attributed to the unique interlocking surface of the nanotube array, consisting of both a global, large scale and a short-range randomness.

Very black infrared detector from vertically aligned carbon nanotubes and electric-field poling of lithium tantalate.

Vertically aligned multiwall carbon nanotubes were grown by water-assisted chemical vapor deposition on a large-area lithium tantalate pyroelectric detector. The processing parameters are nominally identical to those by which others have achieved the "world's darkest substance" on a silicon substrate. The pyroelectric detector material, though a good candidate for such a coating, presents additional challenges and outcomes. After coating, a cycle of heating, electric field poling, and cooling was employed to restore the spontaneous polarization perpendicular to the detector electrodes. The detector responsivity is reported along with imaging as well as visible and infrared reflectance measurements of the detector and a silicon witness sample. We find that the detector responsivity is slightly compromised by the heat of processing and the coating properties are substrate dependent. However, it is possible to achieve nearly ideal values of detector reflectance uniformly less than 0.1% from 400 nm to 4 microm and less than 1% from 4 to 14 microm.