Search NASA⌕ Search

SEARCH · Search NASA

Results for “test correlation”

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

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

At least 1,225 records · Page 68
All sourcesNASA NTRSDOE OSTI
DOE OSTI2016

Complete Status Report Documenting Weld Development for Thin Wall Tubing of ODS Ferritic Alloys

Beginning in 2015, research in the FCRD program began the development of FSW for joining thin sections of 14YWT in the form of thin (0.5 mm) plate and ultimately thin wall tubing. In the previous fiscal year, a ~1 mm thick plate, or sheet, of 14YWT was produced by hot rolling with no edge cracking. The initial FSW experiment was performed on the 1 mm thick plate and involved a bead-on-plate weld in which the spinning pin tool is plunged into the plate surface, but does not penetrate the thickness of the plate, and then travels the length of the plate. The FSW run successfully produced a bead-on-plate stir zone on the 1 mm thick plate of 14YWT, but no characterization studies of the stir zone were performed by the end of FY15. Therefore, the results presented in this report cover the microstructural analysis of the bead-on-plate stir zone and the initial research task on obtaining tensile properties of the stir zone using the digital image correlation (DIC) approach during testing of miniature tensile specimens to assess the quality of the FSW parameters used in the initial experiment. The results of the microstructural characterization study using optical, scanning electron and scanning transmission electron microscopies showed the grain structure in the SZ to have isotropic and irregular shape but very similar size compared to the highly elongated grains oriented horizontally with the plane of the plate that were observed in the unaffected zone of 14YWT. Several cracks oriented horizontally were observed mostly on the retreating side of the SZ in both the SZ and TMAZ. These cracks may have formed due to insufficient pressure being exerted on the top surface of the plate by the shoulder and pin tool during the FSW run. High resolution STEM-EDS analysis showed the presence of the Y-Ti-O particles in the SZ, but that some particles exhibited coarsening. Overall, the FSW parameters used to produce the bead-on-plate SZ in the 0.1 cm thick plate of 14YWT were nearly optimized. The results of the digital image correlation (DIC) analysis of the two SS-Mini-2 tensile specimens fabricated from the 0.1 cm thick plate of 14YWT showed that the specimens exhibited high strength and good ductility. However, strain localization occurred in one of the specimens during the tensile test that was too close to the grips, which invalidated the data from the DIC analysis. This was surprising since the abrupt crack pop-in that occurred in the 0.1 cm thick plate of 14YWT during fabrication by wire EDM suggested that residual stresses were high. Residual stress measurements and the effects of post weld heat treatment on the FSW quality of joined 14YWT plates will be investigated in the next FW work package.

36 MATERIALS SCIENCE↗
NASA NTRS2004

Crash Simulation of a Vertical Drop Test of a B737 Fuselage Section with Overhead Bins and Luggage

The focus of this paper is to describe a crash simulation of a 30-ft/s vertical drop test of a Boeing 737 (B737) fuselage section. The drop test of the 10-ft. long fuselage section of a B737 aircraft was conducted in November of 2000 at the FAA Technical Center in Atlantic City, NJ. The fuselage section was outfitted with two different commercial overhead stowage bins. In addition, 3,229-lbs. of luggage were packed in the cargo hold to represent a maximum take-off weight condition. The main objective of the test was to evaluate the response and failure modes of the overhead stowage bins in a narrow-body transport fuselage section when subjected to a severe, but survivable, impact. A secondary objective of the test was to generate experimental data for correlation with the crash simulation. A full-scale 3-dimensional finite element model of the fuselage section was developed and a crash simulation was conducted using the explicit, nonlinear transient dynamic code, MSC.Dytran. Pre-test predictions of the fuselage and overhead bin responses were generated for correlation with the drop test data. A description of the finite element model and an assessment of the analytical/experimental correlation are presented. In addition, suggestions for modifications to the model to improve correlation are proposed.

Jackson, Karen E.↗
NASA NTRS1989

Flight motor set 360L001 (STS-26R). (Reconstructed dynamic loads analysis)

A transient analysis was performed to correlate the predicted versus measured behavior of the Redesigned Solid Rocket Booster (RSRB) during Flight 360L001 (STS-26R) liftoff. Approximately 9 accelerometers, 152 strain gages, and 104 girth gages were bonded to the motors during this event. Prior to Flight 360L001, a finite element model of the RSRB was analyzed to predict the accelerations, strains, and displacements measured by this developmental flight instrumentation (DFI) within an order of magnitude. Subsequently, an analysis has been performed which uses actual Flight 360L001 liftoff loading conditions, and makes more precise predictions for the RSRB structural behavior. Essential information describing the analytical model, analytical techniques used, correlation of the predicted versus measured RSRB behavior, and conclusions, are presented. A detailed model of the RSRB was developed and correlated for use in analyzing the motor behavior during liftoff loading conditions. This finite element model, referred to as the RSRB global model, uses super-element techniques to model all components of the RSRB. The objective of the RSRB global model is to accurately predict deflections and gap openings in the field joints to an accuracy of approximately 0.001 inch. The model of the field joint component was correlated to Referee and Joint Environment Simulation (JES) tests. The accuracy of the assembled RSRB global model was validated by correlation to static-fire tests such DM-8, DM-9, QM-7, and QM-8. This validated RSRB global model was used to predict RSRB structural behavior and joint gap opening during Flight 360L001 liftoff. The results of a transient analysis of the RSRB global model with imposed liftoff loading conditions are presented. Rockwell used many gage measurements to reconstruct the load parameters which were imposed on the RSRB during the Flight 360L001 liftoff. Each load parameter, and its application, is described. Also presented are conclusions and recommendations based on the analysis of this load case and the resulting correlation between predicted and measured RSRB structural behavior.

Call, V. B.↗
NASA NTRSDate not supplied

Single and Multi-Node Modeling of Direct, Submerged, and Self-Pressurization of A Cryogenic Propellant Tank Using Nodal Tools

The pressurization of cryogenic propellant tanks will always be an important process so long as cryogenic liquids are being considered as fuel sources or used for other in-space applications. Pressure control of the tank ullage is necessary for managing propellant flowrates to an engine or a receiver tank, and modeling of the process is used to predict the pressurant requirements and the amount of propellant boiloff. Direct ullage pressurization is the more traditional approach to tank pressurization, as the physics are straight-forward, and ample test (flight) data have been collected and analyzed over the past several decades. Submerged injection pressurization is an alternate method for tank pressurization and has been shown to reduce pressurant requirements, subcool the propellant, and reduce the risk of ullage collapse. Additionally, the pressurant gas entering the ullage is usually much colder when using the submerged pressurization approach, resulting in reduced propellant boiloff. These benefits are at the expense of vaporizing a small percentage of the propellent. Both tank pressurization methods are viable options for current and future space missions, and it is important to have the capability of analyzing the tank ullage conditions for both approaches. Our previous work has demonstrated the development of a Generalized Fluid System Simulation Program (GFSSP) model, which contains a thermodynamic equilibrium heat and mass transfer subroutine capable of effectively analyzing both direct and submerged pressurization systems [1-2]. This subroutine has most recently been enhanced to include the non-equilibrium effect of pressurant dissolution into the propellant. To date the ullage has always been represented as a single node, and although the simulated single-node temperatures have good comparison with the volume-averaged ullage temperatures computed from test data, the physics of the thermal stratification in the ullage were never captured, and adjustment factors in the model were required. The purpose of this paper is to introduce the development of a multi-node ullage model using GFSSP and to discuss the improvements of the simulated ullage temperature distribution and its resulting effects on ullage heat transfer processes. Test data from the Cryogenic Propellant Storage and Transfer Engineering Developmental Unit (CPST EDU) was used for model validation. For additional comparison, a Thermal Desktop (TD) model was also developed to analyze the CPST EDU direct ullage pressurization tests using both a single node and multi-node approach. The model includes the direct pressurant line, vent line, fill/drain line, and a TD FloCAD Compartment. The TD FloCAD Compartment is employed to represent the liquid and ullage as single volumes inside the tank, to include a liquid/vapor interface, and to generate network level objects such as lumps (analogous to nodes in GFSSP), paths, and ties between the fluid and thermal elements. An established heat load on the model tank was leveraged from a pre-existing higher-fidelity model correlated to CPST EDU test data.

pressurization↗
NASA NTRS1977

Summer performance results obtained from simultaneously testing ten solar collectors outdoors

Ten solar collectors were simultaneously tested outdoors. Efficiency data were correlated using a method that separates solar variables (flux, incident angle) from the desired performance parameters (heat loss, absorbtance, transmittance) which are unique to a given collector design. Tests were conducted on both clear and moderately cloudy days. Correlating data in the above manner, a 2-glass, black paint collector exhibited a decrease in efficiency of 5 percentage points relative to the baseline data for an exposure time of 2 years, 4 months. Condensation on the collector glazing was thought to be a contributing factor in this efficiency change.

Miller, D. R.↗
DOE OSTI2021

Regal Beloit Final Technical Report

The original project proposal submitted from NovaTorque Inc (NovaTorque) in 2016 was to improve the existing motor with 95% efficiency by reducing the losses by 21% to achieve 96%. A few months after the proposal was submitted, NovaTorque lost funding and went out of business. The assets of NovaTorque were then acquired by Regal Beloit Corporation (“Regal Beloit” or “Regal”). When the NovaTorque proposal was selected, the project was transferred to Regal Beloit. Once we had production samples from the new Regal production line, they were tested at the Regal Beloit test lab in Wausau, WI. The original NovaTorque motor had an efficiency of 95%, but when the technology was transferred to Regal Beloit, there were multiple manufacturing improvements made, even though the basic electromagnetic design did not change. The test results in Wausau showed that the motors made at the Regal Beloit plant had an efficiency of 96%. Since the goal of the project was to reduce the losses by 21%, a new efficiency target of 96.8% became the project objective. In the first budget period, motors were tested to get the baseline performance. We then used FEA modeling with ANSYS Maxwell to model the existing motor to get correlation between the FEA simulations and the actual test results. Once the model was validated, we evaluated some changes that could be made to the stator to reduce the losses and improve efficiency without changing the rotor or stator housing, keeping the modifications easy to implement. The changes were primarily in the area of making the stator axially shorter and adding Soft Magnetic Composite (SMC) tooth tips. In the second budget period, we proceeded to design and build the new stator that was identified above and identified additional improvements in the process that included a stator machining modification and the use of rectangular wire. When we actually built the motor, the choice of rectangular wire turned out to be a problem. The wire was made by squishing round wire to get the rectangular shape, which caused work hardening, making the wire too stiff to make the desired coils. We shifted to annealed square wire which was better, but we still could not maintain the proper coil envelope. The result was that we had reduced cross section area for stator laminations. We also had to have a radial offset resulting in a radial misalignment between the stator and rotor because of the oversized coils. After the motor was completed, it was shipped to Texas A&M University for testing. With the loss of flux from these issues, the measured motor efficiency was only 96.3%. The primary focus of this motor was to make sure our FEA simulation model predicted the measured losses and overall efficiency, this we moved on to the FEA simulation. The FEA simulations of the motor “as built” with misalignments had good correlation with the test results, so the next step was to use that model to optimize the design of the motor for a final build. This time we considered changes to the stator and rotor and also minor changes to the housing diameter. In the third budget period, we did the detailed design and construction of the final prototypes. The final prototypes had a slight increase in the stator diameter to fit in a standard Regal Beloit housing. We shortened the stator and use more layers of wire in the coils. We also increased the cone angle of the rotor and stator from the original 110 degrees to 130 degrees to get some additional efficiency and optimized the stator cross section. The predicted efficiency from the FEA simulations was 96.9%. When completed, the motors were tested. We were a little short of reaching our target efficiency goal of 96.8%. We were only able to get to 96.7% efficiency. While it may be possible with additional iterations in designs and future builds to gain that additional 0.1%, we believe that we are close to the best we can achieve from a practical viewpoint, and additional iterations would be more work than the potential gains would be worth. No other motor in this class can even reach the 96% that we started with.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2023

State-of-the-art and review of condensation heat transfer for small modular reactor passive safety: Computational studies

The small modular reactor (SMR) is a promising option with added safety features, economical manufacturing, reliable parts, portability, and scalable energy capacity that emits no greenhouse gas during its operating lifespan. The SMR safety systems, however, need to be evaluated for design and licensing. Thus, they require the verification and validation of the physics models and correlations. This study focuses on state-of-the-art condensation heat transfer analysis and a review of previous studies related to the passive containment cooling system (PCCS) of a SMR. In the PCCS of a SMR, due to its smaller size containment, filmwise condensation is dominant and therefore emphasized in this study. Furthermore, previous condensation heat transfer studies for PCCSs did not make the SMR the primary focus, so a critical review for formulating the state-of-the-art is necessary. A previous review covered experimental condensation heat transfer studies with a brief overview of associated test facilities and empirical correlations. This review covers the empirical, resistance-layer and theoretical (numerical and commercial CFD) approaches.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
NASA NTRS2006

Patterns in Crew-Initiated Photography of Earth from ISS - Is Earth Observation a Salutogenic Experience?

To provide for the well-being of crewmembers on future exploration missions, understanding how space station crewmembers handle the inherently stressful isolation and confinement during long-duration missions is important. A recent retrospective survey of previously flown astronauts found that the most commonly reported psychologically enriching aspects of spaceflight had to do with their Perceptions of Earth. Crewmembers onboard the International Space Station (ISS) photograph Earth through the station windows. Some of these photographs are in response to requests from scientists on the ground through the Crew Earth Observations (CEO) payload. Other photographs taken by crewmembers have not been in response to these formal requests. The automatically recorded data from the camera provides a dataset that can be used to test hypotheses about factors correlated with self-initiated crewmember photography. The present study used objective in-flight data to corroborate the previous questionnaire finding and to further investigate the nature of voluntary Earth-Observation activity. We examined the distribution of photographs with respect to time, crew, and subject matter. We also determined whether the frequency fluctuated in conjunction with major mission events such as vehicle dockings, and extra-vehicular activities (EVAs, or spacewalks), relative to the norm for the relevant crew. We also examined the influence of geographic and temporal patterns on frequency of Earth photography activities. We tested the hypotheses that there would be peak photography intensity over locations of personal interest, and on weekends. From December 2001 through October 2005 (Expeditions 4-11) crewmembers took 144,180 photographs of Earth with time and date automatically recorded by the camera. Of the time-stamped photographs, 84.5% were crew-initiated, and not in response to CEO requests. Preliminary analysis indicated some phasing in patterns of photography during the course of a mission (significant quadratic and trimodal models). There was also a small but significant increase in photo activity on the weekends. In contrast, fewer photos were taken during major station events and for a period of time immediately preceding those events. Data on photography patterns presented here represent a relatively objective group-level measure of Earth observing activities on ISS. Crew Earth Observations offers a self-initiated positive activity that may be important in salutogenesis (maintenance of well-being) of astronauts on long-duration missions. Consideration should be given to developing substitute activities for crewmembers in future exploration missions where there will not be the opportunity to look at Earth, such as on long-duration transits to Mars.

Robinson, Julie A.↗
NASA NTRS1974

Analysis of cross beam runs taken at Haswell Field site, October 1969

A series of cross beam experiments was conducted at the ESSA field site at Haswell, Colorado. Both the optical visual and infrared units were operated on a comparison basis. A data report of the initial auto- and cross-correlation evaluation of these tests is presented. The original test data from the Haswell tests were recorded on F M magnetic tape. A copy of the original tapes was made and employed for the present analysis. The tapes were played back at 60 inches per second for analysis. Some of the infrared units runs (Run Nos. 8,25,40,41,42,43,44,45) were played back at 30 inches per second in order to gain sufficient averaging time. The averaging time was nominally 3200 seconds real time at 60 inches per second and 1600 seconds real time at 30 inches per second. These averaging times correspond to 100 seconds of averaging time of the correlator.

Huff, L. L.↗
NASA NTRS1971

Buckling analysis for axially compressed flat plates, structural sections, and stiffened plates reinforced with laminated composites

A classical buckling analysis is developed for stiffened, flat plates composed of a series of linked plate and beam elements. Plates are idealized as multilayered orthotropic elements. Structural beads and lips are idealized as beams. The loaded edges of the stiffened plate are simply-supported and the conditions at the unloaded edges can be prescribed arbitrarily. The plate and beam elements are matched along their common junctions for displacement continuity and force equilibrium in an exact manner. Offsets between elements are considered in the analysis. Buckling under uniaxial compressive load for plates, sections, and stiffened plates is investigated. Buckling loads are the lowest of all possible general and local failure modes, and the mode shape is used to determine whether buckling is a local or general instability. Numerical correlations with existing analysis and test data for plates, sections, and stiffened plates including boron-reinforced structures are discussed. In general correlations are reasonably good.

Viswanathan, A. V.↗
NASA NTRS1972

Buckling analysis for structural sections and stiffened plates reinforced with laminated composites.

A classical buckling analysis is developed for stiffened, flat plates composed of a series of linked flat plate and beam elements. Plates are idealized as multilayered orthotropic elements; structural beads and lips are idealized as beams. The loaded edges of the stiffened plate are simply supported and the conditions at the unloaded edges can be prescribed arbitrarily. The plate and beam elements are matched along their common junctions for displacement continuity and force equilibrium in an exact manner. Offsets between elements are considered in the analysis. Buckling under uniaxial compressive load for plates, sections and stiffened plates is investigated. Buckling loads are found as the lowest of all possible general and local failure modes and the mode shape is used to determine whether buckling is a local or general instability. Numerical correlations with existing analysis and test data for plates, sections and stiffened plates including boron-reinforced structures are discussed. In general, correlations are reasonably good.

Viswanathan, A. V.↗
NASA NTRS2019

Modal Testing of a Flexible Wing on a Dynamically Active Test Fixture Using Fixed Base Correction Method - IFASD 2019

In modal testing and finite element model correlation, analysts desire modal results using free-free or rigid boundary conditions to ease comparisons of test versus analytical data. It is often expensive both in cost and schedule to build and test with boundary conditions that replicate the free-free or rigid boundaries. Static test fixtures for load testing are often large, heavy, and unyielding, but do not provide adequate boundaries for modal tests because they are dynamically too flexible and often contain natural frequencies within the test article frequency range of interest. The dynamic coupling between the test article and test fixture complicates the model updating process because significant effort needs to be spent on modeling the test fixture and boundary conditions in addition to the test article. If the modal results could be corrected for fixture coupling, then setups used for other structural testing could be adequate for modal testing and would allow significant schedule and cost savings by eliminating a unique setup for only modal testing. To simplify future modal tests, this paper reports on a fixed base correction method that was investigated during modal testing of a full-scale, half-span, flexible wing cantilevered from a static test fixture. The results of this fixed base correction approach look very promising. The method aided in producing similar wing modal characteristics for two different physical boundary configurations of a dynamically active test fixture.

modal testing↗
NASA NTRS2020

Passive Aeroelastic Tailored Wing Modal Test Using the Fixed Base Correction Method

In modal testing and finite element model correlation, analysts desire modal results using free-free or rigid boundary conditions to ease comparisons of test versus analytical data. It is often expensive both in cost and schedule to build and test with boundary conditions that replicate the free-free or rigid boundaries. Static test fixtures for load testing are often large, heavy, and unyielding, but do not provide adequate boundaries for modal tests because they are dynamically too flexible and often contain natural frequencies within the frequency range of interest of the test article. Dynamic coupling between the test article and test fixture complicates the model updating process because significant effort is required to model the test fixture and boundary conditions in addition to the test article. If there were a way to correct the modal results for fixture coupling, then setups used for other structural testing could be adequate for modal testing. In the case described in this paper, a partial static loads testing setup was used, which allowed significant schedule and cost savings by eliminating a unique setup for a modal test. A fixed base correction technique was investigated during modal testing of a flexible wing cantilevered from part of a static test fixture. The technique was successfully used to measure the wing modes de-coupled from the dynamically active test fixture. The technique is promising for future aircraft applications, but more research is needed.

structural dynamics↗
NASA NTRS2021

Modal Testing of a Flexible Wing on a Dynamically Active Test Fixture Using the Fixed Base Correction Method

In modal testing and finite element model correlation, analysts desire modal results using free-free or rigid boundary conditions to ease comparisons of test versus analytical data. It is often expensive both in cost and schedule to build and test with boundary conditions that replicate the free-free or rigid boundaries. Static test fixtures for load testing are often large, heavy, and unyielding, and not provide adequate boundaries for modal tests because they are dynamically too flexible and often contain natural frequencies within the test article frequency range of interest. The dynamic coupling between the test article and test fixture complicates the model updating process because significant effort needs to be spent on modeling the test fixture and boundary conditions in addition to the test article. If the modal results could be corrected for fixture coupling, then setups used for other structural testing could be adequate for modal testing and would allow significant schedule and cost savings by eliminating a unique setup for only modal testing. To simplify future modal tests, this report describes a Fixed Base Correction method that was investigated during modal testing of a full-scale, half-span, flexible wing cantilevered from a static test fixture. The results of this Fixed Base Correction approach look very promising. The method aided in producing similar wing modal characteristics for two different physical boundary configurations of a dynamically active test fixture.

Natalie D Spivey↗
NASA NTRS2023

Measurement of Test Time in the NASA Ames EAST Expansion Cone Facility in Entry 63 and Possible Techniques for Increasing Test Time

The Ames EAST facility and its modes of operation (expansion tube, expansion cone) are described. A buffer section may or may not be used between the driver and the driven tube. Previous expansion tube (without the cone) entries are reviewed. Test times of 4 to 5 microsec were achieved at shock speeds up to 11 km/s. The test conditions and results of the present entry (entry 63) (with the cone) are given. Test times were found to correlate roughly with the density ratio across the buffer gas - driven gas or driver - driven gas interface. Tests with a 12 inch long conical driver without a buffer yielded three runs with test times of ~8 km/s and 3 runs with test times of 4 to 6 microsecs. With a buffer, test times for 4 runs of ~9 microsecs and for 2 runs of 3 to 6 microsecs were obtained. Adding argon to the buffer was found not to improve the test time. Tests with a 30 inch long driver without a buffer yielded test times of 14+, ~10, 7.5 (2 runs) and 5.5 (2 runs) microsecs. Based on one shot, data for the 30 inch driver suggests that test times up to ~14 microseconds could be obtained. The repeatability of that shot, of course, must be established. Suggestions are given to increase test times further.

Electric arc shock tube↗
NASA NTRS1989

Characterization of fluid flow by digital correlation of scattered light

The objective is to produce a physical system suitable for a space environment that can measure fluid velocities in a three-dimensional volume by the development of a particle correlation velocimetry technique. Experimental studies were conducted on a field test cell to demonstrate the suitability and accuracy of digital correlation techniques for measuring two-dimensional fluid flows. This objective was satisfied by: (1) the design of an appropriate illumination and detection system for making velocity measurements within a test cell; (2) the design and construction of a test cell; (3) the preliminary evaluations on fluid and seeding requirements; and (4) the performance of controlled tests using a multiple exposure correlation technique. This presentation is represented by viewgraphs with very little text.

Gilbert, John A.↗
DOE OSTI2019

MaLTESE: Large-Scale Simulation-Driven Machine Learning for Transient Driving Cycles

Optimal engine operation during a transient driving cycle is the key to achieving greater fuel economy, engine efficiency, and reduced emissions. In order to achieve continuously optimal engine operation, engine calibration methods use a combination of static correlations obtained from dynamometer tests for steady-state operating points and road and/or track performance data. As the parameter space of control variables, design variable constraints, and objective functions increases, the cost and duration for optimal calibration become prohibitively large. In order to reduce the number of dynamometer tests required for calibrating modern engines, a large-scale simulation-driven machine learning approach is presented in this work. A parallel, fast, robust, physics-based reduced-order engine simulator is used to obtain performance and emission characteristics of engines over a wide range of control parameters under various transient driving conditions (drive cycles). We scale the simulation up to 3,906 nodes of the Theta supercomputer at the Argonne Leadership Computing Facility to generate data required to train a machine learning model. The trained model is then used to predict various engine parameters of interest, and the results are compared with those predicted by the engine simulator. Our results show that a deepneural-network-based surrogate model achieves high accuracy: Pearson product-moment correlation values larger than 0.99 and mean absolute percentage error within 1.07% for various engine parameters such as exhaust temperature, exhaust pressure, nitric oxide, and engine torque. Once trained, the deep-neural-network-based surrogate model is fast for inference: it requires about 16 mu s for predicting the engine performance and emissions for a single design configuration compared with about 0.5 s per configuration with the engine simulator. Moreover, we demonstrate that transfer learning and retraining can be leveraged to incrementally retrain the surrogate model to cope with new configurations that fall outside the training data space.

deep learning↗