Search NASA⌕ Search

SEARCH · Search NASA

Results for “Normalization”

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 235 records · Page 13

Doping dependence of the magnitude of fluctuating spin moments in the normal state of the pnictide superconductor Sr ( F e 1 - x C o x ) 2 A s 2 inferred from photoemission spectroscopy

We report systematic temperature- and doping-dependent measurements of the Fe3s core-level photoemission spectra in the normal state of superconducting Sr(Fe 1- x Co x ) 2 As 2 . The analysis of the Fe3s spectrum provides an element-specific determination of the mean value of the magnitude of the Fe spin moment measured on the fast (10 -16 -10 -15 s) timescale of the photoemission process. The data reveal the ubiquitous presence in the normal state of Fe spin moments with magnitude fluctuating on short timescales. The data reveal a significant reduction of the magnitude of the effective Fe spin moment on going from the parent to the optimal doped compound. The doping dependence of the magnitude of the spin moment at higher doping level is less clear, being either constant, or even nonmonotonic, depending on temperature. This phenomenology indicates the importance of the interaction between spin and itinerant degrees of freedom in shaping the properties of the normal state. These findings reaffirm the complexity of the normal state of 122 Fe-pnictides, which are typically viewed as the least correlated of the high-temperature unconventional superconductors.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Fast and accurate simulations of calorimeter showers with normalizing flows

In this study, we introduce caloflow, a fast detector simulation framework based on normalizing flows. For the first time, we demonstrate that normalizing flows can reproduce many-channel calorimeter showers with extremely high fidelity, providing a fresh alternative to computationally expensive geant4 simulations, as well as other state-of-the-art fast simulation frameworks based on generative adversarial networks (GANs) or variational autoencoders (VAEs). In addition to the usual histograms of physical features and images of calorimeter showers, we introduce a new metric for judging the quality of generative modeling: the performance of a classifier trained to differentiate real from generated images. We show that GAN-generated images can be identified by the classifier with nearly 100% accuracy, while images generated from caloflow are better able to fool the classifier. More broadly, normalizing flows offer several advantages compared to other state-of-the-art approaches (GANs and VAEs), including tractable likelihoods, stable and convergent training, and principled model selection. Normalizing flows also provide a bijective mapping between data and the latent space, which could have other applications beyond simulation, for example, to detector unfolding.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Normalizing Flows for Microscopic Many-Body Calculations: An Application to the Nuclear Equation of State

We report that normalizing flows are a class of machine learning models used to construct a complex distribution through a bijective mapping of a simple base distribution. We demonstrate that normalizing flows are particularly well suited as a Monte Carlo integration framework for quantum many-body calculations that require the repeated evaluation of high-dimensional integrals across smoothly varying integrands and integration regions. As an example, we consider the finite-temperature nuclear equation of state. An important advantage of normalizing flows is the ability to build highly expressive models of the target integrand, which we demonstrate enables precise evaluations of the nuclear free energy and its derivatives. Furthermore, we show that a normalizing flow model trained on one target integrand can be used to efficiently calculate related integrals when the temperature, density, or nuclear force is varied. This work will support future efforts to build microscopic equations of state for numerical simulations of supernovae and neutron star mergers that employ state-of-the-art nuclear forces and many-body methods.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Exactness of the normal-ordered two-body truncation of three-nucleon forces

Reference-state-based many-body methods start from Hamiltonians that are normal ordered with respect to the reference state. In low-energy nuclear physics applications, normal-ordered Hamiltonians consisting of two- and three-nucleon forces are usually truncated at the two-body rank with residual three-nucleon operators being discarded. Benchmark computations have shown that this truncation is accurate, but we lack an understanding about why it works. Here, we show that the normal-ordered two-body truncation is exact for zero-range three-body forces when nuclei are computed using the coupled cluster with singles and doubles method. As the nuclear three-nucleon force is short ranged and a three-body contact is a leading term in effective field theories of quantum chromodynamics, our result provides an analytical basis for the popular normal-ordered two-body approximation.

Rothman, Maxwell [Univ. of Tennessee, Knoxville, T↗

Structural improvement of unliganded simian immunodeficiency virus gp120 core by normal-mode-based X-ray crystallographic refinement

The structural model of the unliganded and fully glycosylated simian immunodeficiency virus gp120 core determined to 4.0 Å resolution was substantially improved using a recently developed normal-mode-based anisotropic B-factor refinement method. The envelope protein gp120/gp41 of simian and human immunodeficiency viruses plays a critical role in viral entry into host cells. However, the extraordinarily high structural flexibility and heavy glycosylation of the protein have presented enormous difficulties in the pursuit of high-resolution structural investigation of some of its conformational states. An unliganded and fully glycosylated gp120 core structure was recently determined to 4.0 Å resolution. The rather low data-to-parameter ratio limited refinement efforts in the original structure determination. In this work, refinement of this gp120 core structure was carried out using a normal-mode-based refinement method that has been shown in previous studies to be effective in improving models of a supramolecular complex at 3.42 Å resolution and of a membrane protein at 3.2 Å resolution. By using only the first four nonzero lowest-frequency normal modes to construct the anisotropic thermal parameters, combined with manual adjustments and standard positional refinement using REFMAC5, the structural model of the gp120 core was significantly improved in many aspects, including substantial decreases in R factors, better fitting of several flexible regions in electron-density maps, the addition of five new sugar rings at four glycan chains and an excellent correlation of the B-factor distribution with known structural flexibility. These results further underscore the effectiveness of this normal-mode-based method in improving models of protein and nonprotein components in low-resolution X-ray structures.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A hybrid electron and photon IMRT planning technique that lowers normal tissue integral patient dose using standard hardware

Purpose: To present a mixed electron and photon IMRT planning technique using electron beams with an energy range of 6-22 MeV and standard hardware that minimizes integral dose to patients for targets as deep as 7.5 cm. Methods: Ten brain cases, two lung, a thyroid, an abdominal, and a parotid case were planned using two planning techniques: a photon-only IMRT (IMRT) versus a mixed modality treatment (E + IMRT) that includes an enface electron beam and a photon IMRT portion that ensures a uniform target coverage. The electron beam is delivered using a regular cutout placed in an electron cone. The electron energy was chosen to provide a good trade-off between minimizing integral dose and generating a uniform, deliverable plan. The authors choose electron energies that cover the deepest part of PTV with the 65%-70% isodose line. The normal tissue integral dose, the dose for ring structures around the PTV, and the volumes of the 75%, 50%, and 25% isosurfaces were used to compare the dose distributions generated by the two planning techniques. Results: The normal tissue integral dose was lowered by about 20% by the E + IMRT plans compared to the photon-only IMRT ones for most studied cases. With the exception of lungs, the dose reduction associated to the E + IMRT plans was more pronounced further away from the target. The average dose ratio delivered to the 0-2 cm and the 2-4 cm ring structures for brain patients for the two planning techniques were 89.6% and 70.8%, respectively. The enhanced dose sparing away from the target for the brain patients can also be observed in the ratio of the 75%, 50%, and 25% isodose line volumes for the two techniques, which decreases from 85.5% to 72.6% and further to 65.1%, respectively. For lungs, the lateral electron beams used in the E + IMRT plans were perpendicular to the mostly anterior/posterior photon beams, generating much more conformal plans. Conclusions: The authors proved that even using the existing electron delivery hardware, a mixed electron/photon planning technique (E + IMRT) can decrease the normal tissue integral dose compared to a photon-only IMRT plan. Different planning approaches can be enabled by the use of an electron beam directed toward organs at risk distal to the target, which are still spared due the rapid dose fall-off of the electron beam. Examples of such cases are the lateral electron beams in the thoracic region that do not irradiate the heart and contralateral lung, electron beams pointed toward kidneys in the abdominal region, or beams treating brain lesions pointed toward the brainstem or optical apparatus. For brain, electron vertex beams can also be used without irradiating the whole body. Since radiation retreatments become more and more common, minimizing the normal tissue integral dose and the dose delivered to tissues surrounding the target, as enabled by E + IMRT type techniques, should receive more attention.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

SU-E-I-18: CT Scanner QA Using Normalized CTDI Ratio

Purpose: To create a ratio of weighted computed tomography dose index (CTDIw) data normalized to in-air measurements (CTDIair) as a function of beam quality to create a look-up table for frequent, rapid quality assurance (QA) checks of CTDI. Methods: The CTDIw values were measured according to TG-63 protocol using a pencil ionization chamber (Unfors Xi CT detector) and head and body Polymethyl methacrylate (PMMA) phantoms (16 and 32 cm diameter, respectively). Single scan dose profiles were measured at each clinically available energy (80,100,120,140 kVp) on three different CT scanners (two Siemens SOMATOM Definition Flash and one GE Optima), using a tube current of 400 mA, a one second rotation time, and the widest available beam width (32 × 0.6 mm and 16 × 1.25 mm, respectively). These values were normalized to CTDIair measurements using the same conditions as CTDIw. The ratios (expressed in cGy/R) were assessed for each scanner as a function of each energy's half value layer (HVL) paired with the phantom's appropriate bow tie filter measured in mmAl. Results: Normalized CTDI values vary linearly with HVL for both the head and body phantoms. The ratios for the two Siemens machines are very similar at each energy. Compared to the GE scanner, these values vary between 10–20% for each kVp setting. Differences in CTDIair contribute most to the deviation of the ratios across machines. Ratios are independent of both mAs and collimation. Conclusion: Look-up tables constructed of normalized CTDI values as a function of HVL can be used to derive CTDIw data from only three in-air measurements (one for CTDIair and two with added filtration for HVL) to allow for simple, frequent QA checks without CT phantom setup. Future investigations will involve comparing results with Monte Carlo simulations for validation.

07 ISOTOPE AND RADIATION SOURCES↗

SU-E-I-42: Normalized Embryo/fetus Doses for Fluoroscopically Guided Pacemaker Implantation Procedures Calculated Using a Monte Carlo Technique

Purpose: It is well known that pacemaker implantation is sometimes needed in pregnant patients with symptomatic bradycardia. To our knowledge, there is no reported experience regarding radiation doses to the unborn child resulting from fluoroscopy during pacemaker implantation. The purpose of the current study was to develop a method for estimating embryo/fetus dose from fluoroscopically guided pacemaker implantation procedures performed on pregnant patients during all trimesters of gestation. Methods: The Monte Carlo N-Particle (MCNP) radiation transport code was employed in this study. Three mathematical anthropomorphic phantoms representing the average pregnant patient at the first, second and third trimesters of gestation were generated using Bodybuilder software (White Rock science, White Rock, NM). The normalized embryo/fetus dose from the posteroanterior (PA), the 30° left-anterior oblique (LAO) and the 30° right-anterior oblique (RAO) projections were calculated for a wide range of kVp (50–120 kVp) and total filtration values (2.5–9.0 mm Al). Results: The results consist of radiation doses normalized to a) entrance skin dose (ESD) and b) dose area product (DAP) so that the dose to the unborn child from any fluoroscopic technique and x-ray device used can be calculated. ESD normalized doses ranged from 0.008 (PA, first trimester) to 2.519 μGy/mGy (RAO, third trimester). DAP normalized doses ranged from 0.051 (PA, first trimester) to 12.852 μGy/Gycm2 (RAO, third trimester). Conclusion: Embryo/fetus doses from fluoroscopically guided pacemaker implantation procedures performed on pregnant patients during all stages of gestation can be estimated using the method developed in this study. This study was supported by the Greek Ministry of Education and Religious Affairs, General Secretariat for Research and Technology, Operational Program ‘Education and Lifelong Learning’, ARISTIA (Research project: CONCERT)

60 APPLIED LIFE SCIENCES↗

SU-E-J-190: Characterization of Radiation Induced CT Number Changes in Tumor and Normal Lung During Radiation Therapy for Lung Cancer

Purpose: To measure CT number (CTN) changes in tumor and normal lung as a function of radiation therapy (RT) dose during the course of RT delivery for lung cancer using daily IGRT CT images and single respiration phase CT images. Methods: 4D CT acquired during planning simulation and daily 3D CT acquired during daily IGRT for 10 lung cancer cases randomly selected in terms of age, caner type and stage, were analyzed using an in-house developed software tool. All patients were treated in 2 Gy fractions to primary tumors and involved nodal regions. Regions enclosed by a series of isodose surfaces in normal lung were delineated. The obtained contours along with target contours (GTVs) were populated to each singlephase planning CT and daily CT. CTN in term of Hounsfield Unit (HU) of each voxel in these delineated regions were collectively analyzed using histogram, mean, mode and linear correlation. Results: Respiration induced normal lung CTN change, as analyzed from single-phase planning CTs, ranged from 9 to 23 (±2) HU for the patients studied. Normal lung CTN change was as large as 50 (±12) HU over the entire treatment course, was dose and patient dependent and was measurable with dose changes as low as 1.5 Gy. For patients with obvious tumor volume regression, CTN within the GTV drops monotonically as much as 10 (±1) HU during the early fractions with a total dose of 20 Gy delivered. The GTV and CTN reductions are significantly correlated with correlation coefficient >0.95. Conclusion: Significant RT dose induced CTN changes in lung tissue and tumor region can be observed during even the early phase of RT delivery, and may potentially be used for early prediction of radiation response. Single respiration phase CT images have dramatically reduced statistical noise in ROIs, making daily dose response evaluation possible.

60 APPLIED LIFE SCIENCES↗

SU-E-T-630: Predictive Modeling of Mortality, Tumor Control, and Normal Tissue Complications After Stereotactic Body Radiotherapy for Stage I Non-Small Cell Lung Cancer

Purpose: While rates of local control have been well characterized after stereotactic body radiotherapy (SBRT) for stage I non-small cell lung cancer (NSCLC), less data are available characterizing survival and normal tissue toxicities, and no validated models exist assessing these parameters after SBRT. We evaluate the reliability of various machine learning techniques when applied to radiation oncology datasets to create predictive models of mortality, tumor control, and normal tissue complications. Methods: A dataset of 204 consecutive patients with stage I non-small cell lung cancer (NSCLC) treated with stereotactic body radiotherapy (SBRT) at the University of Pennsylvania between 2009 and 2013 was used to create predictive models of tumor control, normal tissue complications, and mortality in this IRB-approved study. Nearly 200 data fields of detailed patient- and tumor-specific information, radiotherapy dosimetric measurements, and clinical outcomes data were collected. Predictive models were created for local tumor control, 1- and 3-year overall survival, and nodal failure using 60% of the data (leaving the remainder as a test set). After applying feature selection and dimensionality reduction, nonlinear support vector classification was applied to the resulting features. Models were evaluated for accuracy and area under ROC curve on the 81-patient test set. Results: Models for common events in the dataset (such as mortality at one year) had the highest predictive power (AUC = .67, p < 0.05). For rare occurrences such as radiation pneumonitis and local failure (each occurring in less than 10% of patients), too few events were present to create reliable models. Conclusion: Although this study demonstrates the validity of predictive analytics using information extracted from patient medical records and can most reliably predict for survival after SBRT, larger sample sizes are needed to develop predictive models for normal tissue toxicities and more advanced machine learning methodologies need be consider in the future.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

MO-F-CAMPUS-T-01: Radiosurgery of Multiple Brain Metastases with Single-Isocenter VMAT: Optimizing Treatment Geometry to Reduce Normal Brain Dose

Purpose: To develop an optimization algorithm to reduce normal brain dose by optimizing couch and collimator angles for single isocenter multiple targets treatment of stereotactic radiosurgery. Methods: Three metastatic brain lesions were retrospectively planned using single-isocenter volumetric modulated arc therapy (VMAT). Three matrices were developed to calculate the projection of each lesion on Beam’s Eye View (BEV) by the rotating couch, collimator and gantry respectively. The island blocking problem was addressed by computing the total area of open space between any two lesions with shared MLC leaf pairs. The couch and collimator angles resulting in the smallest open areas were the optimized angles for each treatment arc. Two treatment plans with and without couch and collimator angle optimization were developed using the same objective functions and to achieve 99% of each target volume receiving full prescription dose of 18Gy. Plan quality was evaluated by calculating each target’s Conformity Index (CI), Gradient Index (GI), and Homogeneity index (HI), and absolute volume of normal brain V8Gy, V10Gy, V12Gy, and V14Gy. Results: Using the new couch/collimator optimization strategy, dose to normal brain tissue was reduced substantially. V8, V10, V12, and V14 decreased by 2.3%, 3.6%, 3.5%, and 6%, respectively. There were no significant differences in the conformity index, gradient index, and homogeneity index between two treatment plans with and without the new optimization algorithm. Conclusion: We have developed a solution to the island blocking problem in delivering radiation to multiple brain metastases with shared isocenter. Significant reduction in dose to normal brain was achieved by using optimal couch and collimator angles that minimize total area of open space between any of the two lesions with shared MLC leaf pairs. This technique has been integrated into Eclipse treatment system using scripting API.

60 APPLIED LIFE SCIENCES↗

A method for normalizing pathology images to improve feature extraction for quantitative pathology

Purpose: With the advent of digital slide scanning technologies and the potential proliferation of large repositories of digital pathology images, many research studies can leverage these data for biomedical discovery and to develop clinical applications. However, quantitative analysis of digital pathology images is impeded by batch effects generated by varied staining protocols and staining conditions of pathological slides. Methods: To overcome this problem, this paper proposes a novel, fully automated stain normalization method to reduce batch effects and thus aid research in digital pathology applications. Their method, intensity centering and histogram equalization (ICHE), normalizes a diverse set of pathology images by first scaling the centroids of the intensity histograms to a common point and then applying a modified version of contrast-limited adaptive histogram equalization. Normalization was performed on two datasets of digitized hematoxylin and eosin (H&E) slides of different tissue slices from the same lung tumor, and one immunohistochemistry dataset of digitized slides created by restaining one of the H&E datasets. Results: The ICHE method was evaluated based on image intensity values, quantitative features, and the effect on downstream applications, such as a computer aided diagnosis. For comparison, three methods from the literature were reimplemented and evaluated using the same criteria. The authors found that ICHE not only improved performance compared with un-normalized images, but in most cases showed improvement compared with previous methods for correcting batch effects in the literature. Conclusions: ICHE may be a useful preprocessing step a digital pathology image processing pipeline.

61 RADIATION PROTECTION AND DOSIMETRY↗

Characterization of photoinduced normal state through charge density wave in superconducting YBa 2 Cu 3 O 6.67

The normal state of high-T c cuprates has been considered one of the essential topics in high-temperature superconductivity research. However, compared to the high magnetic field study of it, understanding a photoinduced normal state remains elusive. Here, we explore a photoinduced normal state of YBa 2 Cu 3 O 6.67 through a charge density wave (CDW) with time-resolved resonant soft x-ray scattering, as well as a high magnetic field x-ray scattering. In the nonequilibrium state where people predict a quenched superconducting state based on the previous optical spectroscopies, we experimentally observed a similar analogy to the competition between superconductivity and CDW shown in the equilibrium state. We further observe that the broken pairing states in the superconducting CuO 2 plane via the optical pump lead to nucleation of three-dimensional CDW precursor correlation. Ultimately, these findings provide a critical clue that the characteristics of the photoinduced normal state show a solid resemblance to those under magnetic fields in equilibrium conditions.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Asymptotic Normalization Coefficients in a Potential Model Involving Forbidden States

It is shown that values obtained for asymptotic normalization coefficients by means of a potential fitted to experimental data on elastic scattering depend substantially on the presence and the number n of possible forbidden states in the fitted potential. The present analysis was performed within exactly solvable potential models for various nuclear systems and various potentials without and with allowance for Coulomb interaction. Various methods for changing the number n that are based on the use of various versions of the change in the parameters of the potential model were studied. A compact analytic expression for the asymptotic normalization coefficients was derived for the case of the Hulthén potential. Specifically, the d + α and α + {sup 12}C systems, which are of importance for astrophysics, were examined. It was concluded that an incorrect choice of n could lead to a substantial errors in determining the asymptotic normalization coefficients. From the results of our calculations, it also follows that, for systems with a low binding energy and, as a consequence, with a large value of the Coulomb parameter, the inclusion of the Coulomb interaction may radically change the asymptotic normalization coefficients, increasing them sharply.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Time-resolved Proteome Profiling of Normal Lung Development

The biochemical networks mediating normal lung development remain unclear with important implications for ameliorating morbidity and mortality in premature infants. Here we perform integrative proteomics and metabolomics analyses of murine lungs from embryonic to early adult ages to characterize the molecular networks mediating normal lung development. Multi-omics analysis identified 8,932 proteins and 158 polar metabolites providing a deep and comprehensive view of the lung proteome and metabolome during normal development. Analysis of the proteomics data revealed ten discrete expression modules driving lung development and the underlying regulatory and signaling network modulating these expression modules. Of interest our data suggest the action of protein ubiquitination in regulating lung development extends beyond the pseudoglandular stage previously described. Combined analysis of the proteomics and metabolomics supported the notion of increased cell proliferation earlier in lung development. Additionally, comparison of the proteomics data with a recent transcriptome analysis revealed biological processes under post-transcriptional control during lung development. Our study provides a unique resource for understanding normal lung development, and potentially avenues for developing strategies to address disrupted development.

computational network analysis, lung development, ↗

The Role of Normal Stress and Shear Stress Heterogeneity in the Inferred Depth-Independence of Stress Drop

Earthquake stress drops are inferred to be independent of their source depth,contradicting standard linear scaling predictions for frictional stick-slip modelsof earthquakes, assuming increasing fault normal stress due to rockoverburden. Here, we examine the scaling between averaged stress drops andincreasing normal stress for simulated earthquakes sequences in continuumrate-and-state fault models. Our models exhibit a power-law-like scaling that isweaker than the linearity predicted by traditional friction models. This resultoccurs as the fault dimension becomes increasingly larger than the earthquakenucleation scale with increasing normal stress. Consequently, the averagedbehavior of ruptures becomes increasingly dominated by conditions for rupturepropagation, reflecting more heterogeneous shear stress conditions. As naturalfaults can be considerably larger than the smallest earthquakes they host, suchweaker scaling between averaged rupture conditions and normal stress maypartially explain the lack of an inferred depth dependence of earthquake stressdrops.

Yang, Minghan↗

Theory and praxis of map analsys in CHEF part 2: Nonlinear normal form

This is the second of three memos describing how normal form map analysis is implemented in CHEF. The first [1] explained the manipulations required to assure that initial, linear transformations preserved Poincare invariants, thereby confirming correct normalization of action-angle coordinates. In this one, the transformation will be extended to nonlinear terms. The third, describing how the algorithms were implemented within the software of CHEF's libraries, most likely will never be written. The first section, Section 2, quickly lays out preliminary concepts and relationships. In Section 3, we shall review the perturbation theory - an iterative sequence of transformations that converts a nonlinear mapping into its normal form - and examine the equation which moves calculations from one step to the next. Following that is a section titled 'Interpretation', which identifies connections between the normalized mappings and idealized, integrable, fictitious Hamiltonian models. A final section contains closing comments, some of which may - but probably will not - preview work to be done later. My reasons for writing this memo and its predecessor have already been expressed. [1] To them can be added this: 'black box code' encourages users to proceed with little or no understanding of what it does or how it operates. So far, CHEF has avoided this trap admirably by failing to attract potential users. However, we reached a watershed last year: even I now have difficulty following the software through its maze of operations. Extensions to CHEF's physics functionalities, software upgrades, and even simple maintenance are becoming more difficult than they should. I hope these memos will mark parts of the maze for easier navigation in the future. Despite appearances to the contrary, I tried to include no (or very little) more than the minimum needed to understand what CHEF's nonlinear analysis modules do.1 As with the first memo, material has been lifted - and modified - from Intermediate Classical Dynamics (ICD) [2], old technical memos, seminar viewgraphs, and lecture notes. Finally, for a reason I do not know but am willing to indulge, equation and comment labels start from where they left off in Part 1.

43 PARTICLE ACCELERATORS↗

Evaluation of Geometrically Nonlinear Reduced Order Models with Nonlinear Normal Modes

Several reduced-order modeling strategies have been developed to create low-order models of geometrically nonlinear structures from detailed finite element models, allowing one to compute the dynamic response of the structure at a dramatically reduced cost. But, the parameters of these reduced-order models are estimated by applying a series of static loads to the finite element model, and the quality of the reduced-order model can be highly sensitive to the amplitudes of the static load cases used and to the type/number of modes used in the basis. Our paper proposes to combine reduced-order modeling and numerical continuation to estimate the nonlinear normal modes of geometrically nonlinear finite element models. Not only does this make it possible to compute the nonlinear normal modes far more quickly than existing approaches, but the nonlinear normal modes are also shown to be an excellent metric by which the quality of the reduced-order model can be assessed. Hence, the second contribution of this work is to demonstrate how nonlinear normal modes can be used as a metric by which nonlinear reduced-order models can be compared. Moreover, various reduced-order models with hardening nonlinearities are compared for two different structures to demonstrate these concepts: a clamped–clamped beam model, and a more complicated finite element model of an exhaust panel cover.

97 MATHEMATICS AND COMPUTING↗