Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “fast optimization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,495 records · Page 83
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Parallel-in-Time Simulation of Lindblad's Equation

Constructing fast quantum logic gates is critical to building a scalable quantum computer. We consider a qudit, a quantum version of a bit that can take an arbitrary number of states, coupled with a cavity. In this project, we wish to force the qudit to reach the 0-state, for any possible initial state. The coupled system changes in time according to Lindblad’s equation, an ordinary differential equation on the density matrix of the quantum system. Lindblad’s equation contains some parameters that we can control, so-called control functions. We seek control functions which force the qudit to the 0-state within 2 microseconds, which is much faster than what is currently done in practice. The search method is gradient descent, a numerical optimization method that uses gradient information to iteratively improve the control parameters. My contribution to this project is an attempt to speed up the computation of the gradient. It currently takes about 40 seconds to compute the gradient which involves solving a set of ODEs sequentially. Current supercomputers have thousands of cores, but sequential computations can only make use of 1 core at a time. We wish to divide up the work better, so that we can use many more cores at once. To this end, we have implemented the Multigrid Reduction in Time (MGRIT) algorithm. We perform a systematic parameter search on how to best apply this algorithm. Results indicate a 25 percent speed up for solving Lindblad’s equation and determining how close the final state is the 0-state.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Fast ignition inertial fusion energy using laser-driven ion beams

Ion fast ignition (IFI), or fusion fast ignition initiated by a laser-driven ion beam, is a promising path to high-gain inertial fusion energy (IFE). In IFI, cold, dense deuterium-tritium (DT) fuel is first assembled using lasers or pulsed power drivers. Then, a high-power ion beam is focused onto a small volume within the fuel (the hot spot), heating the fuel rapidly to conditions where fusion ignition takes place. Fusion burn in this hot spot propagates to the fuel surrounding the hot spot, leading to burnup of a significant fraction of this fuel and the possibility of high gain (G~100), as needed for inertial fusion energy. IFI uses separate drivers for the two basic elements, fuel compression and ignition, allowing maximum control and optimization of each. On the other hand, conventional laser fusion uses multiple beams of the same driver to compress the fuel and shock-heat its very center to ignite a burn wave. Despite impressive progress in conventional laser fusion, the precise spatial symmetry, temporal pulse shaping and timing required for high gain and IFE remain a serious unmet challenge.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2023

Graph Contractions for Calculating Correlation Functions in Lattice QCD

Computing correlation functions for many-particle systems in Lattice QCD is vital to extract nuclear physics observables like the energy spectrum of hadrons such as protons. However, this type of calculation has long been considered to be very challenging and computing-resource intensive because of the complex nature of a hadron composed of quarks with many degrees of freedom. In particular, a correlation function can be calculated through a sum of all possible pairs of quark contractions, each of which is a batched tensor contraction, dictated by Wick's theorem. Because the number of terms of this sum can be very large for any hadronic system of interest, fast evaluation of the sum faces several challenges: an extremely large number of contractions, a huge memory footprint at runtime, and the speed of tensor contractions. In this paper, we present a Lattice QCD analysis software suite, Redstar, which addresses these challenges by utilizing novel algorithmic and software engineering methods targeting modern computing platforms such as many-core CPUs and GPUs. In particular, Redstar represents every term in the sum of a correlation function by a graph, applies efficient graph algorithms to reduce the number of contractions to lower the cost of computations, and minimizes the total memory footprint. Moreover, Redstar carries out the contractions on either CPUs or GPUs utilizing an internal and highly efficient Hadron contraction library. Specifically, we illustrate some important algorithmic optimizations of Redstar, show various key design features of Hadron library, and present the speedup values due to the optimizations along with performance figures for calculating six correlations functions on four computing platforms.

Chen, Jie↗
DOE OSTI2023

Graph Contractions for Calculating Correlation Functions in Lattice QCD

Computing correlation functions for many-particle systems in Lattice QCD is vital to extract nuclear physics observables like the energy spectrum of hadrons such as protons. However, this type of calculation has long been considered to be very challenging and computing-resource intensive because of the complex nature of a hadron composed of quarks with many degrees of freedom. In particular, a correlation function can be calculated through a sum of all possible pairs of quark contractions, each of which is a batched tensor contraction, dictated by Wick's theorem. Because the number of terms of this sum can be very large for any hadronic system of interest, fast evaluation of the sum faces several challenges: an extremely large number of contractions, a huge memory footprint at runtime, and the speed of tensor contractions. In this paper, we present a Lattice QCD analysis software suite, Redstar, which addresses these challenges by utilizing novel algorithmic and software engineering methods targeting modern computing platforms such as many-core CPUs and GPUs. In particular, Redstar represents every term in the sum of a correlation function by a graph, applies efficient graph algorithms to reduce the number of contractions to lower the cost of computations, and minimizes the total memory footprint. Moreover, Redstar carries out the contractions on either CPUs or GPUs utilizing an internal and highly efficient Hadron contraction library. Specifically, we illustrate some important algorithmic optimizations of Redstar, show various key design features of Hadron library, and present the speedup values due to the optimizations along with performance figures for calculating six correlations functions on four computing platforms.

Chen, Jie↗
NASA NTRS1975

Research for preparation of cation-conducting solids by high-pressure synthesis and other methods

It was shown that two body-centered-cubic skeleton structures, the Im3 KSbO3 phase and the defect-pyrochlore phase A(+)B2X6, do exhibit fast Na(+)-ion transport. The placement of anions at the tunnel intersection sites does not impede Na(+)-ion transport in (NaSb)3)(1/6 NaF), and may not in (Na(1+2x)Ta2 5F)(Ox). The activation energies are higher than those found in beta-alumina. There are two possible explanations for the higher activation energy: breathing of the bottleneck (site face or edge) through which the A(+) ions must pass on jumping from one site to another may be easier in a layer structure and/or A(+)-O bonding may be stronger in the cubic structures because the O(2-) ion bonds with two (instead of three) cations of the skeleton. If the former explanation is dominant, a lower activation energy may be achieved by optimizing the lattice parameter. If the latter is dominant, a new structural principle may have to be explored.

Goodenough, J. B.↗
DOE OSTI2024

Optimization of lithium vapor box divertor evaporator location on NSTX-U using SOLPS-ITER

Commercial fusion reactors will be faced with extremely high divertor target heat fluxes that will require mitigation. Simulations of detachment in an NSTX-U scenario projected to have 92 MW/m 2 unmitigated peak target heat flux are presented, which reaches sub-10 MW/m 2 target heat flux using a highly dissipating lithium vapor box divertor design. The lithium vapor box is a detached divertor design which employs lithium vapor evaporation and condensation to contain lithium below the X-point. Previous SOLPS modeling has indicated a lithium vapor box can reduce the heat flux down to 10 MW/m 2 via simultaneous evaporation from the Private Flux Region (PFR) and the Common Flux Region (CFR) sides of the vapor box. It is found here that PFR evaporation has improved access to the separatrix leading to significantly more efficient power dissipation than CFR evaporation. Simulations of target evaporation with an evaporation distribution that is self-consistent with the temperature of a Capillary Porous System with Fast flowing liquid lithium (CPSF) could reach n Li / n e ~ 0.025-0.030 at the Last Closed Flux Surface (LCFS) depending on the liquid metal flow speeds and lithium sputtering yield, while PFR-side evaporation can reach acceptable heat fluxes with n Li / n e ~ 0.038 at the LCFS. However, PFR evaporator performance can be improved if the target is allowed to be hot enough such that it reflects lithium, reaching n Li / n e ~ 0.028 and reducing required lithium evaporation. Ultimately PFR evaporation and target evaporation are found to have similar ability to produce acceptable heat flux solutions with minimal upstream concentration.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Low Cost, High Volume, Carbon Fiber Precursor for Plasma Oxidation

Light weighting with carbon fiber is critical to improving the energy efficiency of wind turbines, airplanes, and automobiles, but the high cost of carbon fiber continues to limit market growth and applications. Half of this cost is in the raw materials. For the carbon fiber market to reach the next level of adoption, the cost of both the raw materials and conversion must be reduced. This project considered the combination of low-cost, textile-grade precursors combined with the increased efficiency of plasma oxidation to greatly reduce the cost of carbon fiber. The primary objective of this project was to make industrial-grade carbon fiber (550 ksi tensile strength, 35 Msi tensile modulus) from textile-grade PAN precursors, something that cannot be achieved with conventional oxidation and to date has not been achieved with any technology. To accomplish this, the project team iteratively optimized the process conditions in the plasma oxidation and carbonization stages to maximize the tensile properties of the resultant carbon fibers. This involved implementing experimental designs for both stages, whereby process conditions (e.g. temperature) were varied and correlated to fiber properties. Samples of oxidized PAN fibers (OPFs) were prepared using a small pilot-scale plasma oxidation oven at 4XTechnologies (4XT). The thermal, chemical, and physical properties of OPF samples were analyzed by the University of Tennessee (UTK) and Oak Ridge National Laboratory (ORNL). The OPF samples were then converted to carbon fiber by subsequent low- and high-temperature carbonization using a lab-scale furnace at ORNL. The resultant carbon fibers were tested for tensile strength also at ORNL, and the results were fed back into the experimental design. Of the hundreds of carbon fiber samples produced over the course of the project, the best sample had an average break strength of 479 ksi and modulus of 33 Msi. While the best tensile properties achieved fell short of the final project targets of 550 ksi for break strength and 35 Msi for modulus, the team demonstrated the potential to produce industrial-grade carbon fiber using textile-grade precursors. The best performing fiber samples had tensile properties that are 87% of the goal for tensile strength and 94% of the goal for tensile modulus. With additional optimization work, it is likely that the targets could be achieved, since only a fraction of the process parameter space was explored during this project. 4XT and its affiliate (4M) have already secured funding through private equity, and there is a CRADA (Cooperative Research and Development Agreement) project currently underway with the Carbon Fiber Technology Center (CFTF) at ORNL to continue this research. Additionally, while not an original goal of the project, the carbon fibers for most samples had diameters >8.5μm. This is considerably larger than commercially available carbon fiber products, which can impart unique properties to the composites (e.g. increased compressive strength) and decrease manufacturing costs even further. Moreover, these fibers were oxidized with plasma oxidation using residence times ranging <60mins, which would be considered fast for normal diameter fibers, but it is as much as 5x faster when factoring in the diameter of the fibers.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Inverse ZrO 2 /Cu as a highly efficient methanol synthesis catalyst from CO 2 hydrogenation

Enhancing the intrinsic activity and space time yield of Cu based heterogeneous methanol synthesis catalysts through CO 2 hydrogenation is one of the major topics in CO 2 conversion into value-added liquid fuels and chemicals. Here we report inverse ZrO 2 /Cu catalysts with a tunable Zr/Cu ratio have been prepared via an oxalate co-precipitation method, showing excellent performance for CO 2 hydrogenation to methanol. Under optimal condition, the catalyst composed by 10% of ZrO 2 supported over 90% of Cu exhibits the highest mass-specific methanol formation rate of 524 g MeOH kg cat -1 h -1 at 220°C, 3.3 times higher than the activity of traditional Cu/ZrO 2 catalysts (159 g MeOH kg cat -1 h -1 ). In situ XRD-PDF, XAFS and AP-XPS structural studies reveal that the inverse ZrO 2 /Cu catalysts are composed of islands of partially reduced 1–2 nm amorphous ZrO 2 supported over metallic Cu particles. The ZrO 2 islands are highly active for the CO 2 activation. Meanwhile, an intermediate of formate adsorbed on the Cu at 1350 cm -1 is discovered by the in situ DRIFTS. This formate intermediate exhibits fast hydrogenation conversion to methoxy. The activation of CO 2 and hydrogenation of all the surface oxygenate intermediates are significantly accelerated over the inverse ZrO 2 /Cu configuration, accounting for the excellent methanol formation activity observed.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Switch Opening and Exchange Method for Stochastic Distribution Network Reconfiguration

The purpose of distribution network reconfiguration (DNR) is to determine the optimal topology of an electricity distribution network, which is an efficient measure to reduce network power losses. Electricity load demand and photovoltaic (PV) output are uncertain and vary with time of day, and will affect the optimal network topology. Single-hour deterministic DNR is incapable of handling this uncertainty and variability. Therefore, this paper proposes to solve a multi-hour stochastic DNR (SDNR). Existing solution methods for DNR are either inaccurate or excessively time-consuming, and are therefore incapable of solving multi-hour SDNRs for large distribution networks. In this regard, a switch opening and exchange (SOE) method is proposed. Starting from a looped network with all switches closed, the SOE consists of three steps. The first step is to sequentially open the switches until all of the loops are opened. The second and third steps modify the status of branches obtained in the first step to obtain better radial topologies. Five test systems are used to validate the accuracy and fast solution speed of the SOE and the superiority of multi-hour SDNR over single-hour deterministic DNR.

distribution network reconfiguration↗
NASA NTRS2000

Laser Brazing of High Temperature Braze Alloy

The Space Shuttle Main Engine (SSME) consists of 1080 conical tubes, which are furnace brazed themselves, manifolds, and surrounding structural jacket making almost four miles of braze joints. Subsequent furnace braze cycles are performed due to localized braze voids between the coolant tubes. SSME nozzle experiences extremely high heat flux (180 mW/sq m) during hot fire. Braze voids between coolant tubes may result in hot combustion gas escape causing jacket bulges. The nozzle can be disqualified for flight or result in mission failure if the braze voids exceed the limits. Localized braze processes were considered to eliminate braze voids, however, damage to the parent materials often prohibited use of such process. Being the only manned flight reusable rocket engine, it has stringent requirement on the braze process. Poor braze quality or damage to the parent materials limits the nozzle service life. The objective of this study was to develop a laser brazing process to provide quality, localized braze joints without adverse affect on the parent materials. Gold (Au-Cu-Ni-Pd-Mn) based high temperature braze alloys were used in both powder and wire form. Thin section iron base superalloy A286 tube was used as substrate materials. Different Laser Systems including CO2 (10.6 micrometers, 1kW), ND:YAG (1.06 micrometers, 4kW). and direct diode laser (808nm. 150W) were investigated for brazing process. The laser process variables including wavelength. laser power, travel speed and angle of inclination were optimized according to bead geometry and braze alloy wetting at minimum heat input level, The properties of laser brazing were compared to that of furnace brazing. Microhardness profiles were used for braze joint property comparison between laser and furnace brazing. The cooling rate of laser brazing was compared to furnace brazing based on secondary dendritic arm spacing, Both optical and Scanning Electron Microscope (SEM) were used to evaluate the microstructures of the braze materials and tube substrate. Metallography of the laser braze joint was compared to the furnace braze. SEM Energy Disperse X-Ray Spectra (EDX) and back scattered imaging were used to analyze braze alloy segregation. Although all of the laser systems, CO2, ND:YAG, and direct diode laser produced good braze joint, the direct diode laser was selected for its system simplicity, compactness and portability. Excellent laser and braze alloy coupling is observed with powder alloy compared to braze alloy wire. Good wetting is found with different gold based braze alloys. The laser brazing process can be optimized so that the adverse affect on the parent materials can be eliminated. Metallography of the laser braze joint has shown that quality braze joint was produced with laser brazing process. Penetration of the laser braze to the substrate is at neglectable level. Zero penetration is observed. Microstructure examinations shown that no observable changes of the microstructure (grain structure and precipitation) in the HAZ area between laser braze and furnace braze. Wide gaps can be laser brazed with single pass for up to 0.024 inches. Finer dendritic structure is observed in laser brazing compared with equiaxial and coarser grain of the furnace brazing microstructure. Greater segregation is also found in the furnace braze. Higher hardness of the laser braze joint comparing to furnace braze is observed due to the fast cooling rate and Finer microstructure in the laser brazing. Laser braze joint properties meet or exceed the furnace joint properties. Direct diode laser for thin section tube brazing with high temperature braze alloys have been successfully demonstrated. The laser's high energy density and precise control has shown significant advantages in reducing process heat input to the substrates and provide high quality braze joints comparing to other localized braze process such as torch, TIG, and MPTA processes. Significant cost savings can be realized particularly with localized braze comparing to a full furnace braze cycle.

Gao, Y. P.↗
DOE OSTI2025

Ion dynamics in hexagonal boron nitride ionogel electrolytes

Ionogel electrolytes incorporating exfoliated hexagonal boron nitride (hBN) nanoplatelets are promising materials for next-generation energy storage systems. However, detailed understanding of their ion transport properties at the molecular level remains limited. This study employs diffusion and relaxation nuclear magnetic resonance (NMR) techniques, including fast-field cycling (FFC) NMR, to investigate the dynamics of ionic species in hBN-ionogels. By spanning a broad frequency range from 30 kHz using FFC NMR to high-field NMR (500–800 MHz), we reveal distinct relaxation mechanisms governing ion dynamics in ionogels with and without lithium salts. Our results highlight the role of hBN in modulating molecular rotation and translational motion, significantly affecting 1 H and 19 F relaxation profiles. The presence of Li + alters the dynamic behavior in ionogels, enhancing anion mobility at the interface. Notably, 7 Li relaxation reveals strong interactions with the hBN surface that cannot be detected by diffusion NMR. Furthermore, these findings underscore the importance of spanning a broad frequency range in NMR studies of ionogels and provide critical insights into optimizing their design as novel electrolytes.

Electrolytes↗
DOE OSTI2020

Computing Bottleneck Structures at Scale for High-Precision Network Performance Analysis

The Theory of Bottleneck Structures is a recently-developed framework for studying the performance of data networks. It describes how local perturbations in one part of the network propagate and interact with others. This framework is a powerful analytical tool that allows network operators to make accurate predictions about network behavior and thereby optimize performance. Previous work implemented a software package for bottleneck structure analysis, but applied it only to toy examples. In this work, we introduce the first software package capable of scaling bottleneck structure analysis to production-size networks. Here, we benchmark our system using logs from ESnet, the Department of Energy's high-performance data network that connects research institutions in the U.S. Using the previously published tool as a baseline, we demonstrate that our system achieves vastly improved performance, constructing the bottleneck structure graphs in 0.21 s and calculating link derivatives in 0.09 s on average. We also study the asymptotic complexity of our core algorithms, demonstrating good scaling properties and strong agreement with theoretical bounds. These results indicate that our new software package can maintain its fast performance when applied to even larger networks. They also show that our software is efficient enough to analyze rapidly changing networks in real time. Overall, we demonstrate the feasibility of applying bottleneck structure analysis to solve practical problems in large, real-world data networks.

benchmark↗
DOE OSTI2022

Energetic particle optimization of quasi-axisymmetric stellarator equilibria

Abstract An important goal of stellarator optimization is to achieve good confinement of energetic particles such as, in the case of a reactor, alphas created by deuterium–tritium fusion. In this work, a fixed-boundary stellarator equilibrium was re-optimized for energetic particle confinement via a two-step process: first, by minimizing deviations from quasi-axisymmetry (QA) on a single flux surface near the mid-radius, and secondly by maintaining this improved QA while minimizing the analytical quantity Γ C , which represents the angle between magnetic flux surfaces and contours of J | | , the second adiabatic invariant. This was performed multiple times, resulting in a group of equilibria with significantly reduced energetic particle losses, as evaluated by Monte Carlo simulations of alpha particles in scaled-up versions of the equilibria. This is the first time that energetic particle losses in a QA stellarator have successfully been reduced by optimizing Γ C . The relationship between energetic particle losses and metrics such as QA error ( E q a ) and Γ C in this set of equilibria were examined via statistical methods and a nearly linear relationship between volume-averaged Γ C and prompt particle losses was found.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS2014

Characterization of MgB2 Superconducting Hot Electron Bolometers

Hot-Electron Bolometer (HEB) mixers have proven to be the best tool for high-resolution spectroscopy at the Terahertz frequencies. However, the current state of the art NbN mixers suffer from a small intermediate frequency (IF) bandwidth as well as a low operating temperature. MgB2 is a promising material for HEB mixer technology in view of its high critical temperature and fast thermal relaxation allowing for a large IF bandwidth. In this work, we have fabricated and characterized thin-film (approximately 15 nanometers) MgB2-based spiral antenna-coupled HEB mixers on SiC substrate. We achieved the IF bandwidth greater than 8 gigahertz at 25 degrees Kelvin and the device noise temperature less than 4000 degrees Kelvin at 9 degrees Kelvin using a 600 gigahertz source. Using temperature dependencies of the radiation power dissipated in the device we have identified the optical loss in the integrated microantenna responsible as a cause of the limited sensitivity of the current mixer devices. From the analysis of the current-voltage (IV) characteristics, we have derived the effective thermal conductance of the mixer device and estimated the required local oscillator power in an optimized device to be approximately 1 microwatts.

Superconducting Devices↗
NASA NTRS2004

A Study on the Requirements for Fast Active Turbine Tip Clearance Control Systems

This paper addresses the requirements of a control system for active turbine tip clearance control in a generic commercial turbofan engine through design and analysis. The control objective is to articulate the shroud in the high pressure turbine section in order to maintain a certain clearance set point given several possible engine transient events. The system must also exhibit reasonable robustness to modeling uncertainties and reasonable noise rejection properties. Two actuators were chosen to fulfill such a requirement, both of which possess different levels of technological readiness: electrohydraulic servovalves and piezoelectric stacks. Identification of design constraints, desired actuator parameters, and actuator limitations are addressed in depth; all of which are intimately tied with the hardware and controller design process. Analytical demonstrations of the performance and robustness characteristics of the two axisymmetric LQG clearance control systems are presented. Takeoff simulation results show that both actuators are capable of maintaining the clearance within acceptable bounds and demonstrate robustness to parameter uncertainty. The present model-based control strategy was employed to demonstrate the tradeoff between performance, control effort, and robustness and to implement optimal state estimation in a noisy engine environment with intent to eliminate ad hoc methods for designing reliable control systems.

DeCastro, Jonathan A.↗
DOE OSTI2023

Data-Driven Affinely Adjustable Robust Volt/VAr Control

Recent years have seen the increasing proliferation of distributed energy resources with intermittent power outputs, posing new challenges to the voltage management in distribution networks. To this end, this paper proposes a data-driven affinely adjustable robust Volt/VAr control (AARVVC) scheme, which modulates the smart inverter’s reactive power in an affine function of its active power, based on the voltage sensitivities with respect to real/reactive power injections. To achieve a fast and accurate estimation of voltage sensitivities, we propose a data-driven method based on deep neural network (DNN), together with a rule-based bus-selection process using the bidirectional search method. Our method only uses the operating statuses of selected buses as inputs to DNN, thus significantly improving the training efficiency and reducing information redundancy. Finally, a distributed consensus-based solution, based on the alternating direction method of multipliers (ADMM), for the AARVVC is applied to decide the inverter’s reactive power adjustment rule with respect to its active power. Only limited information exchange is required between each local agent and the central agent to obtain the slope of the reactive power adjustment rule, and there is no need for the central agent to solve any (sub)optimization problems. Finally, numerical results on the modified IEEE-123 bus system validate the effectiveness and superiority of the proposed data-driven AARVVC method.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS1981

Geologic applications of thermal-inertia mapping from satellite

In the Powder River Basin, Wyo., narrow geologic units having thermal inertias which contrast with their surroundings can be discriminated in optimal images. A few subtle thermal inertia anomalies coincide with areas of helium leakage believed to be associated with deep oil and gas concentrations. The most important results involved delineation of tectonic framework elements some of which were not previously recognized. Thermal and thermal inertia images also permit mapping of geomorphic textural domains. A thermal lineament appears to reveal a basement discontinuity which involves the Homestake Mine in the Black Hill, a zone of Tertiary igneous activity and facies control in oil producing horizons. Applications of these data to the Cabeza Prieta, Ariz., area illustrate their potential for igneous rock type discrimination. Extension to Yellowstone National Park resulted in the detection of additional structural information but surface hydrothermal features could not be distinguished with any confidence. A thermal inertia mapping algorithm, a fast and accurate image registration technique, and an efficient topographic slope and elevation correction method were developed.

Offield, T. W.↗