SEARCH · Search NASA
Results for “Shape from gradient”
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.
Evaluation of INL Supplied MOOSE/OSPREY Model: Modeling Water Adsorption on Type 3A Molecular Sieve
The purpose of this study was to evaluate Idaho National Lab’s Multiphysics Object-Oriented Simulation Environment (MOOSE) software in modeling the adsorption of water onto type 3A molecular sieve (3AMS). MOOSE can be thought-of as a computing framework within which applications modeling specific coupled-phenomena can be developed and run. The application titled Off-gas SeParation and REcoverY (OSPREY) has been developed to model gas sorption in packed columns. The sorbate breakthrough curve calculated by MOOSE/OSPREY was compared to results previously obtained in the deep bed hydration tests conducted at Oak Ridge National Laboratory. The coding framework permits selection of various options, when they exist, for modeling a process. For example, the OSPREY module includes options to model the adsorption equilibrium with a Langmuir model or a generalized statistical thermodynamic adsorption (GSTA) model. The vapor solid equilibria and the operating conditions of the process (e.g., gas phase concentration) are required to calculate the concentration gradient driving the mass transfer between phases. Both the Langmuir and GSTA models were tested in this evaluation. Input variables were either known from experimental conditions, or were available (e.g., density) or were estimated (e.g., thermal conductivity of sorbent) from the literature. Variables were considered independent of time, i.e., rather than having a mass transfer coefficient that varied with time or position in the bed, the parameter was set to remain constant. The calculated results did not coincide with data from laboratory tests. The model accurately estimated the number of bed volumes processed for the given operating parameters, but breakthrough times were not accurately predicted, varying 50% or more from the data. The shape of the breakthrough curves also differed from the experimental data, indicating a much wider sorption band. Model modifications are needed to improve its utility and predictive capability. Recommended improvements include: greater flexibility for input of mass transfer parameters, time-variable gas inlet concentration, direct output of loading and temperature profiles along the bed, and capability to conduct simulations of beds in series.
Counterflow diffusion flames of hydrogen, and hydrogen plus methane, ethylene, propane, and silane vs. air - Strain rates at extinction
Five coaxial tubular opposed jet burners (OJBs) with tube diameter D(T) of 1.8-10 mm and 5 mm conical nozzles were used to form dish-shaped counterflow diffusion flames centered by opposing laminar jets of nitrogen and hydrocarbon-diluted H2 versus air in an argon-purged chamber at 1 atm. Area-averaged air jet velocities at blowoff of the central flame, U(air), characterized extinction of the airside flame as functions of input H2 concentration on the fuelside. A master plot of extensive U(air) data at blowoff versus D(T) shows that U(air) varies linearly with D(T). This and other data sets are used to find that nozzle OJB results for U(air)/diameter average 4.24 + or - 0.28 times larger than tubular OJB results for the same fuel compositions. Critical radial velocity gradients consistent with one-dimensional stagnation point boundary theory and with plug flow inputs are estimated. The results compare favorably with published numerical results based only on potential flow.
Relative Efficiency of TLD-100 to Linear Energy Transfer Radiation: Correction to Astronaut Absorbed Dose
Response of thermoluminescent detectors (TLD-100) to high linear energy transfer (LET) particles has been studied using helium, carbon, silicon, and iron ions from the Heavy Ion Medical Accelerator at Chiba (Japan), iron ions from the Brookhaven National Laboratory (NY) Alternate Gradient Synchrotron, and 53, 134, 185, and 232 MeV protons from the Loma Linda accelerator. Using the measured relative (to (137)Cs dose efficiency, and measured LET spectra from a tissue equivalent proportional counter (TEPC) on 20 Space Shuttle flights, and 7 Mir flights, the underestimation of absorbed dose by these detectors has been evaluated. The dose underestimation is between 15-20% depending upon the flight inclination and shielding location. This has been confirmed by direct correlation of measured dose by TEPC and TLD-100 at a low shielded location in the Shuttle mid-deck. A comparison of efficiency- LET data with a compilation of similar data from TLD-700, shows that shapes of the two curves are nearly identical, but that the TLD-100 curve is systematically lower by about 13%, and is the major cause of dose underestimation. These results strongly suggest that TLDs used for crew dose estimation be regularly calibrated using heavy ions.
Comparison of kinetic and extended magnetohydrodynamics computational models for the linear ion temperature gradient instability in slab geometry
We perform linear stability studies of the ion temperature gradient (ITG) instability in unsheared slab geometry using kinetic and extended magnetohydrodynamics (MHD) models, in the regime k{sub ∥}/k{sub ⊥}≪1. The ITG is a parallel (to B) sound wave that may be destabilized by finite ion Larmor radius (FLR) effects in the presence of a gradient in the equilibrium ion temperature. The ITG is stable in both ideal and resistive MHD; for a given temperature scale length L{sub Ti0}, instability requires that either k{sub ⊥}ρ{sub i} or ρ{sub i}/L{sub Ti0} be sufficiently large. Kinetic models capture FLR effects to all orders in either parameter. In the extended MHD model, these effects are captured only to lowest order by means of the Braginskii ion gyro-viscous stress tensor and the ion diamagnetic heat flux. We present the linear electrostatic dispersion relations for the ITG for both kinetic Vlasov and extended MHD (two-fluid) models in the local approximation. In the low frequency fluid regime, these reduce to the same cubic equation for the complex eigenvalue ω=ω{sub r}+iγ. An explicit solution is derived for the growth rate and real frequency in this regime. These are found to depend on a single non-dimensional parameter. We also compute the eigenvalues and the eigenfunctions with the extended MHD code NIMROD, and a hybrid kinetic δf code that assumes six-dimensional Vlasov ions and isothermal fluid electrons, as functions of k{sub ⊥}ρ{sub i} and ρ{sub i}/L{sub Ti0} using a spatially dependent equilibrium. These solutions are compared with each other, and with the predictions of the local kinetic and fluid dispersion relations. Kinetic and fluid calculations agree well at and near the marginal stability point, but diverge as k{sub ⊥}ρ{sub i} or ρ{sub i}/L{sub Ti0} increases. There is good qualitative agreement between the models for the shape of the unstable global eigenfunction for L{sub Ti0}/ρ{sub i}=30 and 20. The results quantify how far fluid calculations can be extended accurately into the kinetic regime. We conclude that for the linear ITG problem in slab geometry with unsheared magnetic field when k{sub ∥}/k{sub ⊥}≪1, the extended MHD model may be a reliable physical model for this problem when ρ{sub i}/L{sub Ti0}<10{sup −2} and k{sub ⊥}ρ{sub i}<0.2.
Direct numerical simulation of hypersonic turbulent boundary layers: effect of spatial evolution and Reynolds number
Direct numerical simulations (DNS) are performed to investigate the spatial evolution of flat-plate zero-pressure-gradient turbulent boundary layers over long streamwise domains ( >300δ i , with δ i the inflow boundary-layer thickness) at three different Mach numbers, 2.5, 4.9 and 10.9, with the surface temperatures ranging from quasiadiabatic to highly cooled conditions. The settlement of turbulence statistics into a fully developed equilibrium state of the turbulent boundary layer has been carefully monitored, either based on the satisfaction of the von Kármán integral equation or by comparing runs with different inflow turbulence generation techniques. The generated DNS database is used to characterize the streamwise evolution of multiple important variables in the high-Mach-number, cold-wall regime, including the skin friction, the Reynolds analogy factor, the shape factor, the Reynolds stresses, and the fluctuating wall quantities. The data confirm the validity of many classic and newer compressibility transformations at moderately high Reynolds numbers (up to friction Reynolds number Re τ ≈1200 ) and show that, with proper scaling, the sizes of the near-wall streaks and superstructures are insensitive to the Mach number and wall cooling conditions. The strong wall cooling in the hypersonic cold-wall case is found to cause a significant increase in the size of the near-wall turbulence eddies (relative to the boundary-layer thickness), which leads to a reduced-scale separation between the large and small turbulence scales, and in turn to a lack of an outer peak in the spanwise spectra of the streamwise velocity in the logarithmic region.
High-J CO SLEDs in nearby infrared bright galaxies observed by Herschel/PACS
We report the detection of far-infrared (FIR) CO rotational emission from nearby active galactic nuclei (AGNs) and starburst galaxies, as well as several merging systems and Ultra-Luminous Infrared Galaxies (ULIRGs). Using the Herschel Photodetector Array Camera and Spectrometer (PACS), we have detected transitions in the J{sub upp} = 14–30 range. The PACS CO data obtained here provide the first reference of well-sampled FIR extragalactic CO spectral line energy distributions (SLEDs) for this range. We find a large range in the overall SLED shape, even among galaxies of similar type, demonstrating the uncertainties in relying solely on high-J CO diagnostics to characterize the excitation source of a galaxy. Combining our data with low-J line intensities taken from the literature, we present a CO ratio–ratio diagram and discuss its value in distinguishing excitation sources and physical properties of the molecular gas. The position of a galaxy on such a diagram is less a signature of its excitation mechanism, than an indicator of the presence of warm, dense molecular gas. We then quantitatively analyze the CO emission from a subset of the detected sources with single-component and two-component large velocity gradient (LVG) radiative transfer models to fit the CO SLEDs. From these fits we derive the molecular gas mass and the corresponding CO-to-H{sub 2} conversion factor, α{sub CO}, for each respective source. For the ULIRGs we find α values in the canonical range 0.4– 5M{sub ⊙} (K km s{sup −1} pc{sup 2}){sup −1}, while for the other objects, α varies between 0.2 and 14. Finally, we compare our best-fit LVG model results with previous studies of the same galaxies and comment on any differences.
Quantum Mechanical Simulations of Complex Nanostructures for Photovoltaic Applications
A quantitative understanding of the electronic excitations in nanostructures, especially complex nanostructures, is crucial for making new-generation photovoltaic (PV) cells based on nanotechnology, which have high efficiency and low cost. Yet current quantum mechanical simulation methods are either computationally too expensive or not accurate and reliable enough, hindering the rational design of the nanoscale PV cells. The PI seeks to develop new methodologies to overcome the challenges in this very difficult and long-lasting problem, pushing the field forward so that electronic excitations can be accurately predicted for systems involving thousands of atoms. The primary objective of this project is to develop new approaches for electronic excitation calculations that are more accurate than traditional density functional theory (DFT) and are applicable to systems larger than what current beyond-DFT methods can treat. In this proposal, the PI will first address the excited-state problem within the DFT framework to obtain quasiparticle energies from both Kohn-Sham (KS) eigenvalues and orbitals; and the electron-hole binding energy will be computed based on screened Coulomb interaction of corresponding DFT orbitals. The accuracy of these approaches will be examined against many-body methods of GW/BSE and quantum Monte Carlo (QMC). The PI will also work on improving the accuracy and efficiency of the GW/BSE and QMC methods in electronic excitation computations by using better KS orbitals obtained from orbital-dependent DFT as inputs. Then an extended QMC database of ground- and excited-state properties will be developed, and this will be spot checked and supplemented with data from GW/BSE calculations. The investigation will subsequently focus on the development of an improved exchange-correlation (XC) density functional beyond the current generalized gradient approximation (GGA) level of parameterization, with parameters fitted to the QMC database. This will allow the ground-state properties of focus systems to be more precisely predicted using DFT. These new developments will then be applied to investigate a chosen set of complex nanostructures that have great potential for opening new routes in designing materials with improved transport, electronic, and optical properties for PV and other optoelectronic usages: (1) Hybrid interfaces between materials with distinct electronic and optical properties, such as organic molecules (conjugated polymers, e.g. P3HT) and inorganic semiconducting materials (Si and ZnO). Complicated interface structures, including interface bonding configurations, compositional and geometrical blending patterns, interfacial defects, and various sizes and shapes of inorganic nanomaterials, will be considered for the purpose of understanding the working mechanisms of present organic/nano PV systems and designing optimum interface structures for fast charge separation and injection. (2) Complex-structured semiconducting nanomaterials that could induce charge separation without pn- or hetero-junctions. The new methodology will allow the PI to investigate the performance of realistic semiconducting nanomaterials of internal (impurities, defects, etc.) and external (uneven surface, mechanical twisting and bending, surface chemistry, etc.) complexities on optical absorption and charge transport against charge trapping and recombination. Of particular interest is whether such structural complexity in a single material could even be beneficial for PV usage, for example, charge separation through morphology control. Successful completion of the proposed DFT methodology would have a far-reaching impact on our ability to study and exploit the nature of electronic excitations in complex materials, advancing the design of next-generation electronic and optoelectronic devices in all facets of renewable energy conversion and storage, including photovoltaics, thermoelectricity, photochemistry, etc.
Two-dimensional model of a Space Station Freedom thermal energy storage canister
The Solar Dynamic Power Module being developed for Space Station Freedom uses a eutectic mixture of LiF-CaF2 phase change salt contained in toroidal canisters for thermal energy storage. Results are presented from heat transfer analyses of the phase-change salt containment canister. A 2-D, axisymmetric finite-difference computer program which models the canister walls, salt, void, and heat engine working fluid coolant was developed. Analyses included effects of conduction in canister walls and solid salt, conduction and free convection in liquid salt, conduction and radiation across salt vapor filled void regions, and forced convection in the heat engine working fluid. Void shape, location, and growth or shrinkage (due to density difference between the solid and liquid salt phases) were prescribed based on engineering judgement. The salt phase change process was modeled using the enthalpy method. Discussion of results focuses on the role of free-convection in the liquid salt on canister heat transfer performance. This role is shown to be important for interpreting the relationship between groundbased canister performance (in 1-g) and expected on-orbit performance (in micro-g). Attention is also focused on the influence of void heat transfer on canister wall temperature distributions. The large thermal resistance of void regions is shown to accentuate canister hot spots and temperature gradients.
Offshore application of landslide susceptibility mapping using gradient-boosted decision trees: a Gulf of Mexico case study
Abstract Among natural hazards occurring offshore, submarine landslides pose a significant risk to offshore infrastructure installations attached to the seafloor. With the offshore being important for current and future energy production, there is a need to anticipate where future landslide events are likely to occur to support planning and development projects. Using the northern Gulf of Mexico (GoM) as a case study, this paper performs Landslide Susceptibility Mapping (LSM) using a gradient-boosted decision tree (GBDT) model to characterize the spatial patterns of submarine landslide probability over the United States Exclusive Economic Zone (EEZ) where water depths are greater than 120 m. With known spatial extents of historic submarine landslides and a Geographic Information System (GIS) database of known topographical, geomorphological, geological, and geochemical factors, the resulting model was capable of accurately forecasting potential locations of sediment instability. Results of a permutation modelling approach indicated that LSM accuracy is sensitive to the number of unique training locations with model accuracy becoming more stable as the number of training regions was increased. The influence that each input feature had on predicting landslide susceptibility was evaluated using the SHapely Additive exPlanations (SHAP) feature attribution method. Areas of high and very high susceptibility were associated with steep terrain including salt basins and escarpments. This case study serves as an initial assessment of the machine learning (ML) capabilities for producing accurate submarine landslide susceptibility maps given the current state of available natural hazard-related datasets and conveys both successes and limitations.
A discharge with a magnetic X-point as a negative hydrogen ion source
The study presents first results from investigations of a novel low-pressure plasma source, intended for a negative hydrogen ion production. The source utilizes a dc magnetic field, shaped to form a cusp with a magnetic null-point (X-point). Beside the common role of filtering out the high energy electrons, this magnetic field configuration ensures in the present case also an interesting mechanism of coupling the RF power to the plasma. Investigations performed using radio frequency modulation spectroscopy (RFMOS) reveal that the main power coupling to the electrons is confined in the region on one side of the X-point. The modulation of the light intensity indicates also the presence of a strong dc drift close to the plane of the X-point. Several hypothesises for its explanation are raised: an azimuthal diamagnetic drift due to strong axial gradients of the electron energy, the excitation of a standing helicon wave, which couples to the radial magnetic field in the plane of the X-point, or a Trivelpiece-Gould wave which is resonantly absorbed near the plane of the X-point.
Shivalik Plasma Device-I, a glow discharge device to study the collective dynamics of dusty plasma
Using a uniquely configured glow discharge-based Shivalik Plasma Device-I, we demonstrate a variety of collective phenomena in dusty plasma away from the glow discharge region. The cylindrical glass device produces plasma using parallel disc-shaped electrodes with a smaller anode size than the cathode. The dust microparticles are initially sprinkled over the grounded cathode. These particles acquire a significant negative charge upon plasma formation, resulting in their levitation due to the balance between the Coulomb force and gravity. The new device supports the levitation of a big-sized (10 × 8 × 5 cm 3 ) three-dimensional dust cloud over the glass surface. It contrasts the dusty plasma formations in-between electrodes reported earlier. As the discharge voltage varies from high to low, the dust cloud travels from over the glass surface to between the electrodes. A complex interplay of dust void over the cathode, a sharp density gradient, and gravity lead to self-excitation of collective dust phenomena. It includes dust density waves (phase velocity, v ph ~4 cm/s), dust cloud oscillation (frequency, f = 5 Hz), sheared flow (flow velocity, v f ~ 1 cm/s), and multiple-sized dust vortices. These dust vortices provided an excellent platform for studying turbulent mixing phenomena. The power spectrum analysis agreed with two-dimensional Kolmogorov power-law scaling. This is an ideal dusty plasma apparatus where we can create or move the dust cloud to a location of choice from the glass surface to in-between the electrodes and excite one among many collective dust dynamics.
Extending XACC for Quantum Optimal Control
Quantum computing vendors are beginning to open up application programming interfaces for direct pulse-level quantum control. With this, programmers can begin to describe quantum kernels of execution via sequences of arbitrary pulse shapes. This opens new avenues of research and development with regards to smart quantum compilation routines that enable direct translation of higher-level digital assembly representations to these native pulse instructions. In this work, we present an extension to the XACC system-level quantum-classical software framework that directly enables this compilation lowering phase via user-specified quantum optimal control techniques. This extension enables the translation of digital quantum circuit representations to equivalent pulse sequences that are optimal with respect to the backend system dynamics. Our work is modular and extensible, enabling third party optimal control techniques and strategies in both C++ and Python. We demonstrate this extension with familiar gradient-based methods like gradient ascent pulse engineering (GRAPE), gradient optimization of analytic controls (GOAT), and Krotov's method. Our work serves as a foundational component of future quantum-classical compiler designs that lower high-level programmatic representations to low-level machine instructions.
The Path to High Q-Factors in Superconducting Accelerating Cavities: Flux Expulsion and Surface Resistance Optimization
Accelerating cavities are devices resonating in the radio-frequency (RF) range used to accelerate charged particles in accelerators. Superconducting accelerating cavities are made out of niobium and operate at the liquid helium temperature. Even if superconducting, these resonating structures have some RF driven surface resistance that causes power dissipation. In order to decrease as much as possible the power losses, the cavity quality factor must be increased by decreasing the surface resistance. In this dissertation, the RF surface resistance is analyzed for a large variety of cavities made with different state-of-the-art surface treatments, with the goal of finding the surface treatment capable to return the highest Q-factor values in a cryomodule-like environment. This study analyzes not only the superconducting properties described by the BCS surface resistance, which is the contribution that takes into account dissipation due to quasi-particle excitations, but also the increasing of the surface resistance due to trapped flux. When cavities are cooled down below their critical temperature inside a cryomodule, there is always some remnant magnetic field that may be trapped increasing the global RF surface resistance. This thesis also analyzes how the fraction of external magnetic field, which is actually trapped in the cavity during the cooldown, can be minimized. This study is performed on an elliptical single-cell horizontally cooled cavity, resembling the geometry of cavities cooled in accelerator cryomodules. The horizontal cooldown study reveals that, as in case of the vertical cooldown, when the cooling is performed fast, large thermal gradients are created along the cavity helping magnetic flux expulsion. However, for this geometry the complete magnetic flux expulsion from the cavity equator is more difficult to achieve. This becomes even more challenging in presence of orthogonal magnetic field, that is easily trapped on top of the cavity equator causing temperature rising. The physics behind the magnetic flux expulsion is also analyzed, showing that during a fast cooldown the magnetic field structures, called vortices, tend to move in the same direction of the thermal gradient, from the Meissner state region to the mixed state region, minimizing the Gibbs free energy. On the other hand, during a slow cool down, not only the vortices movement is limited by the absence of thermal gradients, but, also, at the end of the superconducting transition, the magnetic field concentrates along randomly distributed normal-conducting region from which it cannot be expelled anymore. The systematic study of the surface resistance components performed for the different surface treatments, reveals that the BCS surface resistance and the trapped flux surface resistance have opposite trends as a function of the surface impurity content, defined by the mean free path. At medium field value, the BCS surface resistance is minimized for nitrogen-doped cavities and significantly larger for standard niobium cavities. On the other hand, Nitrogen-doped cavities show larger dissipation due to trapped flux. This is consequence of the bell-shaped trend of the trapped flux sensitivity as a function of the mean free path. Such experimental findings allow also a better understanding of the RF dissipation due to trapped flux. The best compromise between all the surface resistance components, taking into account the possibility of trapping some external magnetic field, is given by light nitrogen-doping treatments. However, the beneficial effects of the nitrogen-doping is completely lost when large amount of magnetic field is trapped during the cooldown, underlying the importance of both cooldown and magnetic field shielding optimization in high quality factors cryomodules.
Prediction of in-depth gap heating ratios from wing glove model test data
In-depth gap heating ratios were predicted down RSI tile sidewalls based on temperature measurements obtained from the JSC arc-jet Wing Glove model tests in order to develop gap heating ratios which resulted in the best possible fit of test data and to produce a set of engineering verification heating ratios similar in shape to one another which could be used at various body points on the Orbiter during reentry. The Rockwell TPS Multidimensional heat conduction program was used to perform 3-D thermal analyses using a 3.0 in. thick section of a curved RSI tile with 283 nodal points. Correlation with test data shows that the predicted heating ratios were significantly higher down in the gap than the zero pressure values for T/C stacks 39 and 38 on the Wing Glove model. For stack 37 (in a low pressure region), the baseline heating ratio overpredicted the temperature data. This analysis, which showed that the heating ratios were a strong function of the product of pressure and pressure gradient, will be used to compare with recent Gap/Step and Ames Double Wedge test/analysis results in the effort to identify the Orbiter gap response to high delta P flight environment.
Two-dimensional model of a Space Station Freedom thermal energy storage canister
The Solar Dynamic Power Module being developed for Space Station Freedom uses a eutectic mixture of LiF-CaF2 phase change salt contained in toroidal canisters for thermal energy storage. Results are presented from heat transfer analyses of the phase change salt containment canister. A 2-D, axisymmetric finite difference computer program which models the canister walls, salt, void, and heat engine working fluid coolant was developed. Analyses included effects of conduction in canister walls and solid salt, conduction and free convection in liquid salt, conduction and radiation across salt vapor filled void regions and forced convection in the heat engine working fluid. Void shape, location, growth or shrinkage (due to density difference between the solid and liquid salt phases) were prescribed based on engineering judgement. The salt phase change process was modeled using the enthalpy method. Discussion of results focuses on the role of free-convection in the liquid salt on canister heat transfer performance. This role is shown to be important for interpreting the relationship between ground based canister performance (in l-g) and expected on-orbit performance (in micro-g). Attention is also focused on the influence of void heat transfer on canister wall temperature distributions. The large thermal resistance of void regions is shown to accentuate canister hot spots and temperature gradients.
Velocity profile shapes in Alcator C-Mod plasmas
Abstract Toroidal rotation velocity spatial profiles ( r / a < 0.8) have been obtained from C-Mod over a wide range of operational conditions, including H-mode, I-mode, ICRF-heated L-mode and Ohmic L-mode (LOC and SOC), and in plasmas with ITBs, LH wave injection and MCFD. Peaked, flat and hollow rotation profiles have been observed. In H- and I-mode plasmas, generally with co-current peaked profiles, the peaking is correlated with temperature profile peaking, and both increase with toroidal magnetic field (decrease with ρ ∗ ). Any dependence on density peaking is unclear. For Ohmic L-mode discharges, with LOC, the velocity profiles are usually flat and most often directed co-current, while with SOC the profiles are hollow, mostly co-current at the edge and counter-current in the core. Both of these Ohmic rotation states exist with matched density and temperature profiles (and gradients), indicating that neither gradient is relevant during rotation reversals. For plasmas with LH wave injection and discharges with ITBs, the velocity profiles are hollow while the density and temperature profiles exhibit substantial peaking. Broadly speaking for all operational regimes, there is no unifying ordering of the velocity gradient with plasma parameters.
Low–severity fire as a mechanism of organic matter protection in global peatlands: Thermal alteration slows decomposition
Worldwide, regularly-recurring wildfires shape many peatland ecosystems to the extent that fire-adapted species often dominate plant communities, suggesting that wildfire is an integral part of peatland ecology rather than an anomaly. The most destructive blazes are smoldering fires that are usually initiated in periods of drought and can combust entire peatland carbon stores. However, peatland wildfires more typically occur as low-severity surface-burns that arise in the dormant season when vegetation is desiccated, and soil moisture is high. In such low-severity fires, surface layers experience flash heating, but there is little loss of underlying peat to combustion. This study examines the potential importance of such processes in several peatlands that span a gradient from hemiboreal to tropical ecozones and experience a wide range of fire return intervals. We show that low-severity fires can increase the pool of stable soil carbon by thermally altering the chemistry of soil organic matter (SOM), thereby reducing rates of microbial respiration. Using XPS and FTIR, we demonstrate that low-severity fires significantly increase the degree of carbon-condensation and aromatization of SOM functional groups, particularly on the surface of peat aggregates. Laboratory incubations show lower CO 2 emissions from peat subjected to low-severity fire, and predict lower cumulative CO 2 emissions from burned peat after one to three years. Also, low-severity fires reduce the temperature sensitivity (Q 10 ) of peat, indicating that these fires can inhibit microbial access to SOM. The increased stability of thermally-altered SOM may allow a greater proportion of organic matter to survive vertical migration into saturated and anaerobic zones of peatlands where environmental conditions physiochemically protect carbon stores from decomposition for thousands of years. Furthermore, across latitudes, low-severity fire is an overlooked factor influencing carbon cycling in peatlands, which is relevant to global carbon budgets as climate-change alters fire regimes worldwide.