UCLA Civil and Environmental Engineering

Architected materials design across orders of magnitude length scale intrigues exceptional mechanical responses nonexistent in their natural bulk state. However, the so-termed mechanical metamaterials, when scaling bottom down to the atomistic or microparticle level, remain largely unexplored and conventionally fall out of their coarse-resolution, ordered-pattern design space. Here, combining high-throughput molecular dynamics (MD) simulations and machine learning (ML) strategies, some intriguing atomistic families of disordered mechanical metamaterials are discovered, as fabricated by melt quenching and exemplified herein by lightweight-yet-stiff cellular materials featuring a theoretical limit of linear stiffness-density scaling, whose structural disorder-rather than order-is key to reduce the scaling exponent and is simply controlled by the bonding interactions and their directionality that enable flexible tunability experimentally. Importantly, a systematic navigation in the forcefield landscape reveals that, in-between directional and non-directional bonding such as covalent and ionic bonds, modest bond directionality is most likely to promotes disordered packing of polyhedral, stretching-dominated structures responsible for the formation of metamaterials. This work pioneers a bottom-down atomistic scheme to design mechanical metamaterials formatted disorderly, unlocking a largely untapped field in leveraging structural disorder in devising metamaterials atomistically and, potentially, generic to conventional upscaled designs.

As part of the next generation liquefaction (NGL) project, we are developing probabilistic triggering and manifestation models using laboratory data and cone penetration test (CPT) case histories in the NGL database. The case histories are used to develop probabilistic models for surface manifestation conditional on susceptibility, liquefaction triggering, soil properties, stratigraphic details, and other features. Susceptibility is interpreted as a sole function of soil composition and is expressed as a probabilistic function of soil behavior type index, Ic, obtained from CPT. A triggering model is derived based on laboratory tests on high-quality specimens from literature; this model captures mean responses and uncertainty reflective of data dispersion and is considered as a Bayesian prior that will subsequently be updated by field observation data. A manifestation model is then regressed from field case histories where surface manifestation was or was not observed, information on soil conditions that enables identification of layers likely to liquefy, and ground shaking conditions. We describe the approach applied to develop our manifestation model; for a given layer this model considers layer depth, thickness, CPT tip resistance, and Ic. The result of this process is a logistic function in which manifestation probability decreases with increasing depth, decreasing thickness, increasing tip resistance, and increasing Ic. Profile manifestation is then derived by aggregating individual layer manifestation probabilities.

The pore structures of hardened Portland/slag cement pastes (>75 wt% slag content), and the initial capillary absorption of moisture through these pores, were monitored using ex situ synchrotron X-ray computerised microtomography and in situ quantitative neutron radiography. The pore structure becomes more constricted as the cement hydrates and its microstructure develops. This mechanism was effective even at a slag content as high as 90 wt% in the cementitious blend, where the lowest total porosity and a significant pore refinement were identified at extended curing ages (360 d). By combining this information with neutron radiographic imaging, and directly quantifying both depth and mass of water uptake, it was observed that 90 wt% slag cement outperformed the 75 wt% slag blend at 90 days in terms of resistance to capillary water uptake, although the higher-slag blend had not yet developed such a refined microstructure at 28 days of curing. The assumptions associated with the sharp front model for water ingress do not hold true for highly substituted slag cement pastes. Testing transport properties at 28 days may not give a true indication of the performance of these materials in service in the long term.

Layered double hydroxide (LDH) phases that form during cement hydration can incorporate a variety of interlayer anions in their interlayer positions. Here, a range of phases of general formula [MII(1−x)MIII(x)(OH)2][An−]x/n·zH2O were synthesized, where MII = Mg2+ (hydrotalcite) or Ca2+ (AFm), MIII = Al3+ such that [MII/Al] = 2 (Ca and Mg, atomic units) or 3 (Mg only), and A = Cl−, Br−, or I−. All the synthesized phases were characterized to assess their composition, density, and crystal structure. By approach from undersaturation, the solubility data of these compounds was measured at 5, 25, and 60°C. This thermochemical data was used to successfully model their formation using thermodynamic modeling and to infer the fields of stability of these compounds for conditions of relevance to cementitious systems. It is seen that halide-containing hydrotalcite phases strongly compete with hydroxide-containing hydrotalcite, with the latter prevailing at high pH. In contrast, halide-containing AFm compounds are more stable compared with hydroxide-containing AFm compositions.

The NGA-Subduction Project is a multi-year, multidisciplinary project with the goal of developing a ground motion database and ground motion models for global subduction zone earthquakes including those in Japan, Taiwan, Cascadia, Alaska, New Zealand, South America, and Central America. Our ground motion model development is currently at the stage of identifying regional trends in path terms. We use a combination of data inspection and regression techniques to distinguish path effects in the data, including differences between interface and inslab events, forearc/back-arc effects, regional effects, and azimuthal effects. Our approach to model development is to first develop a path model capturing these effects, then to investigate source and site effects. The parameterization of functional form is guided in part by the scaling expected by a generic equivalent point-source stochastic model. We expect regionalization in path and will investigate further regionalization in site response and in overall model bias.

Between August and November 2016, three major earthquake events occurred in Central Italy. The first event, with M6.1, took place on 24 August 2016, the second (M5.9) on 26 October, and the third (M6.5) on 30 October 2016. As part of the Italy-US GEER team investigation, we recorded the amplitude and character of offset on the Mount Vettore Fault Zone (MVFZ) using traditional manual field recording and mapping techniques and advanced state-of-the-art geomatics methods of LIDAR and Structure From Motion. Extensive field surveys by INGV geologists and the GEER team were performed on the flanks of Mt Vettore after the 24 August and 30 October events, and a limited survey was done between the two October events by INGV. These surveys indicated normal offset on several strands of the MVFZ, along the upper flanks of Mount Vettore and on the Piano Grande basin floor. The primary trace of the fault had measurable offset up to 215 cm in the northern section of the fault (42.810N-42.818N), and lesser offsets in the southern and central portion of the fault (42.796N-42.810N). In tandem with the traditional field recording of offset, we collected TLS-LIDAR at several locations and flew approximately 5 km of the fault with unmanned aerial systems (UAS) to image the offsets. Lidar and Structure-from-Motion point cloud models were merged to construct a virtual topographic model of the fault. Comparison between the virtual offsets in the point cloud data and the field measurements at the same location found close agreement within 20% of the measured field values. The results indicate that LIDAR and UAS-based methods for collecting and analyzing topographic fault offsets are accurate and potentially greatly improve the magnitude of fault offset data sets from events with measurable surface rupture.

Zero-emission vehicle (ZEV) adoption is a key climate mitigation tool, but its environmental justice implications remain unclear. Here, we quantify ZEV adoption at the census tract level in California from 2015 to 2020 and project it to 2035 when all new passenger vehicles sold are expected to be ZEVs. We then apply an integrated traffic model together with a dispersion model to simulate air quality changes near roads in the Greater Los Angeles. We found that per capita ZEV ownership in non-disadvantaged communities (non-DACs) as defined by the state of California is 3.8 times of that in DACs. Racial and ethnic minorities owned fewer ZEVs regardless of DAC designation. While DAC residents receive 40% more pollutant reduction than non-DACs due to intercommunity ZEV trips in 2020, they remain disproportionately exposed to higher levels of traffic-related air pollution. With more ZEVs in 2035, the exposure disparity narrows. However, to further reduce disparities, the focus must include trucks, emphasizing the need for targeted ZEV policies that address persistent pollution burdens among DAC and racial and ethnic minority residents.