Search NASA⌕ Search

SEARCH · Search NASA

Results for “spectrogram”

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.

18 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Improving Robustness of Spectrogram Classifiers with Neural Stochastic Differential Equations

Signal analysis and classification is fraught with high levels of noise and perturbation. Computer-vision-based deep learning models applied to spectrograms have proven useful in the field of signal classification and detection; however, these methods aren't designed to handle the low signal-to-noise ratios inherent within non-vision signal processing tasks. While they are powerful, they are currently not the method of choice in the inherently noisy and dynamic critical infrastructure domain, such as smart-grid sensing, anomaly detection, and non-intrusive load monitoring. Currently, these models can be brittle, which makes them susceptible to noisy input. This also means they have sub-optimal stability of explanation outputs. Experts and technicians using these models to make decisions in real world scenarios need assurance that a model is performing as it is supposed to. The classification or prediction outputs it generates should be sound and grounded, not likely to change in the presence of shifting noise landscapes. In this work, we explore the idea of Neural Stochastic Differential Equations (NSDE's) to improve the robustness of models trained to classify time series data and the effect of NSDE's on the explainability of outputs. We then test the effectiveness of these approaches by applying them to a non-intrusive load monitoring (NILM) dataset that consists of simulated harmonic signals injected into a real building.

Brogan, Joel↗
DOE OSTI2025

Embedded Sensing in Additive Manufacturing Metal and Polymer Parts: A Comparative Study of Integration Techniques and Structural Health Monitoring Performance

This study presents a comparative evaluation of post-process sensor integration in additively manufactured (AM) metal and the in-situ process for polymer structures for structural health monitoring (SHM), with an emphasis on embedded sensors. Geometrically identical specimens were fabricated using copper via metal fused filament fabrication (FFF) and PLA via polymer FFF, with piezoelectric transducers (PZTs) inserted into internal cavities to assess the influence of material and placement on sensing fidelity. Mechanical testing under compressive and point loads generated signals that were transformed into time–frequency spectrograms using a Short-Time Fourier Transform (STFT) framework. An engineered RGB representation was developed, combining global amplitude scaling with an amplitude-envelope encoding to enhance contrast and highlight subtle wave features. These spectrograms served as inputs to convolutional neural networks (CNNs) for classification of load conditions and detection of damage-related features. Results showed reliable recognition in both copper and PLA specimens, with CNN classification accuracies exceeding 95%. Embedded PZTs were especially effective in PLA, where signal damping and environmental sensitivity often hinder surface-mounted sensors. This work demonstrates the advantages of embedded sensing in AM structures, particularly when paired with spectrogram-based feature engineering and CNN modeling, advancing real-time SHM for aerospace, energy, and defense applications.

additive manufacturing↗
DOE OSTI2026

A non-intrusive framework using acoustic signals and deep learning for boiling diagnostics in visual-limited environments

Accurate monitoring of boiling heat transfer is critical for safeguarding high-power systems operating in environments where conventional optical diagnostics are hindered by radiation fields or restricted visual accessibility. This study presents a non-intrusive framework that integrates hydroacoustic sensing with deep learning to infer near-wall boiling characteristics and enable predictive thermal assessment without visual access. In a prototypical subcooled flow-boiling facility representative of the Isotope Production Facility (IPF) at Los Alamos, hydrophones capture boiling-induced acoustic emissions that are transformed into background-removed Short-Time Fourier Transform (STFT) spectrograms. A convolutional neural network (CNN) then regresses heat flux, wall superheat, and key bubble parameters directly from these spectrograms. The CNN achieved predictive accuracy under nominal conditions and demonstrated robustness and generalization under acoustic noise for Signal-to-Noise Ratios (SNRs) down to approximately 0 dB. When integrated into an ANSYS CFX wall-boiling model, the acoustically inferred parameters reproduced boiling curve and critical heat flux (CHF) values consistent with image-based benchmarks. Furthermore, the model retained reliable performance under moderate variations in bulk temperature, flow rate, and hydrophone placement, confirming its generalizability across practical boundary conditions. These results demonstrate the feasibility of hydroacoustic-based deep learning as a viable path toward real-time, radiation-tolerant boiling diagnostics and predictive thermal safety assessment in inaccessible systems such as the IPF.

42 ENGINEERING↗
DOE OSTI2025

A Bayesian approach to time-domain photonic Doppler velocimetry analysis

Photonic Doppler velocimetry (PDV) is an established technique for measuring the velocities of fast-moving surfaces in high-energy-density experiments. In the standard approach to PDV analysis, the short-time Fourier transform (STFT) is used to generate a spectrogram from which the velocity history of the target is inferred. The user chooses the form, duration, and separation of the window function. Here, in this study, we present a Bayesian approach to infer the velocity directly from the PDV oscilloscope trace, without using the spectrogram for analysis. This is clearly a difficult inference problem due to the highly periodic nature of the data, but we find that with carefully chosen prior distributions for the model parameters, we can accurately recover the injected velocity from synthetic data. We validate this method using PDV data collected at the STAR two-stage light gas gun at Sandia National Laboratories, recovering shock-front velocities in quartz that are consistent with those inferred using the STFT-based approach and are interpolated across regions of low signal-to-noise data. Although this method does not rely on the same user choices as the STFT, we caution that it can be prone to misspecification if the chosen model is not sufficient to capture the velocity behavior. Analysis using posterior predictive checks can be used to establish whether a better model is required, although more complex models come with additional computational cost, often taking more than several hours to converge when sampling the Bayesian posterior. We, therefore, recommend it be viewed as a complementary method to that of the STFT-based approach.

Allison, James R. [First Light Fusion Ltd., Yarnto↗
DOE OSTI2026

Velocity Extraction Using Complete Time-Domain Waveform Data and Audio Machine Learning

We developed a new machine learning-based tool for extracting information from interferometry measurements: MIDWAZE (Modular Interferometry Direct Waveform AnalyZEr). This paper showcases MIDWAZE’s ability to extract an object’s velocity information from Photonic Doppler Velocimetry (PDV) data at near-human accuracy with little to no human intervention. MIDWAZE can extract velocities roughly 350 times as fast as a human analyst "rushing" to complete their extractions, with similar extraction accuracy. MIDWAZE’s most outstanding feature is that it operates directly in waveform/temporal space, freeing analysis from certain limitations imposed by traditional spectrogram-based approaches and opening the way to "phase aware" PDV analysis. MIDWAZE also has limited ability to discriminate between different solid objects, which we develop as a first step towards automated discrimination of different kinds of objects such as ejecta clouds.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2025

Cyclic Background Noise Variations on Infrasound Microbarometers From Micrometeorology and Human Activity

Infrasound microbarometers deployed on the Earth's surface capture pressure fluctuations and acoustic signatures, revealing trends in surface wind speed and cycles in ambient sound. Previous studies investigated wind noise trends at quarter day resolution and urban acoustic background variations on hourly to weekly scales. Cyclic variations at sub-hourly resolution over local distances remain unaddressed. We show that topography-driven micrometeorology impacts diurnal background noise levels below 10 Hz. Anthropogenic noise occurs over daily and weekly cycles, with inputs from traffic, ventilation systems, and power lines. These noise patterns vary at stations spaced less than a kilometer apart. We observe these phenomena by using a circular spectrogram to visualize frequency trends over a periodic temporal scale. This study demonstrates that infrasound microbarometer deployments can highlight localized wind patterns and time scales of human activity. The results suggest that future microbarometer deployments may benefit from site noise surveys before selecting final sensor locations.

Malach, Amrit Kaur [Sandia National Lab. (SNL-CA),↗
DOE OSTI2026

Learning earthquake ground motions via conditional generative modeling

Predicting high-fidelity ground motions for future earthquakes is crucial for seismic hazard assessment and infrastructure resilience. Conventional empirical simulations suffer from sparse sensor distribution and geographically localized earthquake locations, while physics-based methods are computationally intensive and require accurate representations of Earth structures and earthquake sources. We propose an artificial intelligence (AI) spectrogram generator, Conditional Generative Modeling for Ground Motion (CGM-GM). CGM-GM leverages earthquake magnitudes and geographic coordinates of earthquakes and sensors as inputs, when postprocessed with phase information, capturing spatially continuous Fourier amplitude spectra (FAS) as well as properties such as P and S arrivals, and waveform durations, without explicit physics constraints. This is achieved through a probabilistic autoencoder that extracts latent distributions in the time-frequency domain and variational sequential models for prior and posterior distributions. We evaluate the performance of CGM-GM using small-magnitude earthquake records from the San Francisco Bay Area, a region with high seismic risks. Here, we report that CGM-GM demonstrates potential for complementing physics-based simulations and non-ergodic empirical ground motion models, as well as shows promise in seismology and beyond.

geophysics↗
DOE OSTI2025

Pedestal formation via different trajectories in the stability space in response to the timing scan of neutral beam heating in DIII-D

The frequency of type-I ELMs decreases as the initiation of the neutral beam injection (NBI) heating is delayed with respect to the time when plasma current (I p ) reaches flat-top in the ITER Baseline Scenario discharges in DIII-D. Henceforth, the time gap between the NBI initiation and I p flat-top will be referred to as “heating delay.” As the heating delay is modified, pedestal formation follows different trajectories in the edge current density–pedestal pressure gradient (j edge -∇p e ped ) space from the L-H transition toward the first ELM event. During the stationary phase after the first ELM, the ELM frequency (f ELM ) decreases by a factor of ~2 as the heating delay is increased. A longer pedestal recovery time in the inter-ELM period is observed for the low f ELM discharges as compared to the high f ELM discharges. Both low and high f ELM discharges show nearly identical profiles of electron density and temperature and have a similar MHD stability just before an ELM crash. However, a marked difference is observed in the magnetic spectrogram of the high and low f ELM discharges in response to the variation in the heating delay. The main difference is in the 200–400 kHz range of the magnetic spectra. A quasi-coherent mode (QCM) at 220 kHz and weaker broadband fluctuations are observed in the high f ELM discharges, while only strong broadband fluctuations are prevalent in the low f ELM discharges. ELM-synchronized analysis shows that the time evolution of these modes is different for the high and low f ELM discharges. The localization of both these modes is confirmed at the maximum gradient region of the pedestal. We hypothesize that these modes cause important pedestal transport and that the difference in the pedestal recovery of the high and low f ELM discharges is a result of the difference in transport driven by these modes, as they change with changes in the heating delay. It is demonstrated experimentally for the first time that discharges with similar pedestal parameters can carry the history of the heating delay into the stationary phase and that changes in turbulent-driven transport are a likely cause of changes in f ELM observed with variations of heating delay.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

In situ characterization of two unknown ultrashort laser pulses using four-wave mixing in gas

Accurate characterization of two ultrashort laser pulses is of great interest in many ultrafast pump-probe experiments. We demonstrate a method based on four-wave mixing (FWM) in a gas which could be easily implemented into many existing pump-probe setups with minimal modifications for accurate, in situ characterization of both unknown pulses. This technique is tested on pairs of unknown pulses at wavelengths of 400/800 nm, and 266/400 nm. We measured the spectrogram of the pulse generated through FWM of the two unknown pulses by scanning the delay between two unknown pulses. The retrieval algorithm converges to accurately predict the intensity and the phase profiles of both unknown pulses with a trace error of < 1% and the accuracy is verified using an independent pulse characterization device.

Nambu, Noa (ORCID:0000000197303400)↗
DOE OSTI2025

Denoising Seismograms in the Time Domain Using a Deep Learning Model

Deep learning has emerged as a transformative tool for enhancing the extraction of reliable information from seismograms, addressing the increasing demand for precise and efficient seismic data analysis. We introduce an innovative encoder–decoder deep learning model, named WaveDenoiser, designed for noise reduction in the time domain, thereby eliminating the need for spectrogram computations that have been used for existing deep learning tools and significantly improving processing speed. Utilizing the benchmark dataset that is Stanford Earthquake Dataset, we developed three models of varying sizes: base, medium, and large. Notably, the large (referred to as WaveDenoiser) model demonstrated superior performance, achieving a median signal‐to‐noise ratio improvement of 8.8 dB on in‐distribution unseen data (in the same geographic region) and 7.7 dB on out‐distribution unseen data (in a new geographic region), outpacing both the base and medium models. Further evaluation of the WaveDenoiser model revealed a reduction in median arrival‐time errors by 0.02 s for P waves and 0.01 s for S waves when processing waveforms prior to phase picking using PhaseNet on in‐distribution unseen data. When tested on out‐distribution unseen data, the model also effectively reduced the P‐wave median arrival‐time error by 0.02 and 0.01 s in median arrival‐time error for S waves. Importantly, the application of WaveDenoiser resulted in a significant reduction of phase picking outliers by 1.1% to 3.6% for both P and S waves. In addition, we achieved over five times acceleration in processing speed compared with the seisBench implementation of DeepDenoiser. Our findings underscore the potential of WaveDenoiser as a powerful tool for improving seismic data analysis and processing efficiency.

P-waves↗
DOE OSTI2025

Towards AI Based Data Classification for Decision Making During Testing

During the development of high-consequence items, test systems should be capable of differentiating between test failures resulting from narrowly missing requirements versus those indicating potentially catastrophic faults. In many instances, classifying the data corresponds to simply identifying whether measured waveforms have approximately the anticipated shape. Cast in this light, the problem reduces to converting raw data into a form optimal for use with neural network classifiers. This manuscript investigates different means of representing raw data for image classification. Raw data plots and Short Time Fourier Transform (STFT) spectrograms are classified by both custom built, small-scale, Convolution Neural Networks (CNN) and open-source, multi-million parameter, pre-trained deep CNNs. In the case of time varying frequency content, the STFTs provide images with greater detail and can be accurately classified with simpler networks. This requires less memory and runs faster than classifying the raw data using the more sophisticated options—making STFTs optimal for applications with memory constraints. STFTs are not a panacea. In some cases the time-domain signal contains useful information that should not be discarded. Rather than using raw data or STFTs, the images can be constructed from both by using red and green channels of an RGB image to visualize the real and imaginary components of the transform, with the raw data occupying the blue channel.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Toroidal Alfven Wave Coupling (Nonlinear Wave-Wave Interactions) on DIII-D

Connect DIII-D physics with space plasma phenomena. In this case of using the toroidal Alfvén eigenmodes and frequency-chirping Reversed-Shear Alfvén eignmodes in DIII-D, we will document how the nonlinear interactions among dipolar Kinetic Alfvén Wave eigenmodes in space plasmas may determine saturation levels of these fluctuations. We seek evidence of nonlinear energy transfer and wave-wave coupling during 3-wave interactions mediated by a much lower-frequency mode. In FY2019, we found evidence of nonlinear “wave-wave” interactions in 175 relevant shots of archival DIII-D data. Toroidal mode number was identified and spectrograms were produced from each shot’s Mirnov coil data. Bispectral analysis was performed using a preliminary version of a new user friendly code derived from a 1995 M.S. thesis at WVU. These results formed a part of a May 2019 M.S. thesis at WVU.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

Improving Chirped Fiber Bragg Grating Resolution for Position-Sensitive Sensors in Shock- and Detonation-Driven Experiments

Chirped fiber Bragg gratings (CFBGs) are robust diagnostic sensors that are widely used to track detonation-driven and shock wave propagation. CFBGs are inscribed with a linearly chirped periodic index of refraction changes that alter the Bragg wavelength along the length of the probe. The light return of each individual Bragg element is captured by a detector at a unique time to map the full reflected spectrum. The CFBG spectrum is measured with a dispersive Fourier transform of the reflected light that temporally stretches the spectrum to increase spatial resolution and make a one-to-one map of the wavelength on a time axis. Here, we propose an improvement of CFBG temporal resolution by incorporating two co-linear laser pulses with orthogonal polarization states and a 5 ns time offset. The two separate signals were split and tracked by two separate detectors. An oscilloscope captured good separation in the signals, and two separate spectrograms were generated and interleaved in the post-processing of the data. This novel technique doubled the CFBG temporal resolution and led to a doubled location resolution. As a proof-of-concept of this technique, the resolution improvement was compared between standard CFBG measurements and the two polarization states method on a position-sensitive CFBG sensor. CFBG resolution doubling will advance sensor capabilities and will have a direct impact on improving capture and analysis in dynamic, high-explosive experiments.

42 ENGINEERING↗
DOE OSTI2026

Advancing Industry 4.0: Multimodal Sensor Fusion for AI-Based Fault Detection in 3D Printing

Additive manufacturing, particularly fused deposition modeling, is transforming modern production by enabling rapid prototyping and complex part fabrication. However, its layer-by-layer process remains vulnerable to faults such as nozzle clogging, filament runout, and layer misalignment, which compromise print quality and reliability. Traditional inspection methods are costly, time-intensive, and often limited to post-process analysis, making them unsuitable for real-time intervention. In this current study, the authors developed a novel, low-cost, and portable faultdetection system that leverages multimodal sensor fusion and artificial intelligence for real-time monitoring in FDM-based 3D printing. The system integrates acoustic, vibration, and thermal sensing into a non-intrusive architecture, capturing complementary data streams that reflect both mechanical and process-related anomalies. Acoustic and thermal sensors operate in a fully contactless manner, while the vibration sensor requires minimal attachment such that it will not interfere with printer hardware, thereby preserving portability and ease of deployment. The multimodal signals are processed into spectrograms and time-frequency features, which are classified using convolutional neural networks for intelligent fault detection. The proposed system advances Industry 4.0 objectives by offering an affordable, scalable, and practical monitoring solution that improves faultdetection accuracy, reduces waste, and supports sustainable, adaptive manufacturing.

42 ENGINEERING↗
DOE OSTI2025

Propagation of ion cyclotron emission in the DIII-D tokamak

Experiments in L- and H-mode plasmas performed on the DIII-D tokamak explored ion cyclotron emission (ICE) propagation via the recently upgraded ICE diagnostic. The distance between the plasma and the outer wall was scanned to alter the evanescent region, which was seen to impact the detection of runaway electron-driven whistler modes in a comparable frequency range to typical ICE harmonics (f≳10fci) [Heidbrink et al., Plasma Phys. Controlled Fusion 61, 014007 (2018)]. In L-mode plasmas, core ICE did not exhibit clear trends as the plasma-wall distance was altered. Instead, inadvertent changes to the fast ion density correlated with different ICE fine structures; the most spectral bands (spaced ∼150–200 kHz apart) were observed at the highest fast ion density, and a just single band when nfast decreased by ∼ 40%. The full-wave Petra-M code simulated core ICE propagation through the plasma and to ICE probes, agreeing with experimental findings that these are likely compressional fast waves, which propagate to probe locations with no dependence on plasma-wall gap. Edge ICE in H-mode plasmas was similarly unaffected by plasma-wall distance. Unlike in core ICE cases, there were negligible changes to the edge fast ion distribution and edge bulk plasma profiles. However, changes in ELM frequency seemed most closely tied to ICE behavior. Ultimately, core and edge ICE did not exhibit clear dependencies on the plasma-wall distance, and ICE harmonics were observed with even the largest plasma-wall gaps. This independence bodes well for similar diagnostics in future tokamaks, which might be placed further from high radiation areas without adversely impacting measurement capabilities.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Efficient generation and extreme compression of multidimensional solitary states in molecular gas-filled hollow-core fibers driven by picosecond Yb lasers

We present an in-depth study on the impact of spatiotemporal Raman enhancement in molecular gas-filled hollow-core fibers (HCFs), demonstrating the efficient generation and post-compression of multidimensional solitary states (MDSS). Through different experimental scenarios—employing large-core HCFs filled with molecular gases (N 2 and N 2 O) and driven by high energy, sub-picosecond and picosecond Fourier transform-limited ytterbium laser pulses—this work leverages multimode propagation and enhanced spatiotemporal interactions to achieve significant spectral broadening and asymmetric redshift, contrasting sharply with self-phase modulation. Our findings reveal that, beyond the regime of maximum nonadiabatic molecular alignment, spatiotemporal nonlinear enhancement primarily governs spectral broadening for input pulse durations up to 1 ps. The process shows limited sensitivity to input pulse duration and the two investigated molecular gases (N 2 and N 2 O), with only subtle differences in broadening arising from their distinct Raman spectroscopic properties. Furthermore, post-compression of MDSS was achieved in various cases. Notably, using 7 mJ, 1 ps laser pulses, we generated 22 fs pulses with a 47% energy conversion efficiency of the input pulse energy. These results position MDSS as a powerful platform for generating high-energy, ultrashort pulses with tunable wavelengths, offering a robust solution for applications such as high harmonic generation.

47 OTHER INSTRUMENTATION↗
DOE OSTI2025

Analysis of the HAE activity in the TJ-II stellarator using a Landau closure model

The aim of this study is to analyze the stability of helical Alfvén eigenmodes (HAEs) in TJ-II discharges and the stabilizing effect of the energetic particles generated by the neutral beam injector (NBI) on pressure gradient-driven modes (PGDMs). HAE and PGDM stability is studied using the linear version of the gyro-fluid code FAR3d and the continuous structure by the STELLGAP code. First, Alfvén eigenmode (AE) and PGDM activity observed in the experiments is reproduced by the simulations, identifying unstable m/n = 4/7 − 2/3 and 7/12 − 5/8 HAEs triggered around ρ = 0.66 showing a frequency of 209 and 204 kHz, respectively, as well as 5/3 PGDM. Next, a parametric study is performed with respect to the thermal ion density and iota profile in the middle-outer plasma region to verify the robustness of the simulation results with respect to the uncertainty of experimental profiles. The analysis confirms that experimental uncertainty does not cause large deviations in the simulation results, showing the destabilization of the same HAEs for all the configurations tested. The simulations also indicate the decay of the 5/3 PGDM growth rate as the energetic particle (EP) population in the plasma increases, consistent with the experiment. Stability analysis of the n = 3, 7, 11, n = 5, 9, 13, n = 6, 10, 14, and n = 8, 12 helical families is performed with respect to the NBI operational regime for different EP energies, β as well as deposition profiles. The most unstable configuration is the radially localized on-axis NBI operation (stiff EP density profile gradients nearby the magnetic axis). Using the simulation model that reproduces the observed Alfvén activity, we extend the study to analyze NBI performance within a theoretical framework. It shows that increasing NBI voltage (which raises EP energy) leads to a degradation in NBI performance for a given power (related to EP β and their density). To achieve better NBI operation, higher voltage must be balanced with lower injection power, ensuring stable AEs while keeping the same EP β.

Plasma confinement↗
DOE OSTI2025

Phase-based velocity extraction method for photonic Doppler velocimetry with potential higher time resolution

We present an extension of the [Takeda et al., J. Opt. Soc. Am. 72, 156 (1982)] phase extraction method to heterodyne photonic Doppler velocimetry applications. The method yields results equivalent to those obtained by the short-time Fourier transform (STFT), while offering potential improvements in time resolution. Unlike STFT, which relies on window functions, such as the Hamming window, that emphasize central data points and diminish the influence of edges, the extended Takeda method utilizes all data uniformly. This uniform treatment allows for the derivation of empirical equations that directly relate velocity error to the actual time resolution rather than to the local analysis duration. The established equation provides a useful metric for both optimizing hardware configuration and guiding data analysis. Simulation and experimental results confirm that, for a given dataset, specifying a target time resolution yields consistent velocity errors for both methods. These findings underscore the Takeda method’s advantages, particularly its potential higher time resolution and reduced computational burden, making it a valuable tool for high-throughput applications such as laser dynamic compression experiments.

Computer simulation↗