Transportation noise assessment

Abstract

To provide a method, a system, and a program for assessing noise generated by noise from transportation means (e.g., an eVTOL) under operating conditions.SOLUTION: A method comprises: defining a computer-based model of transportation means; automatically determining aerodynamic performance and propulsive performance of the transportation means based on the model; performing flight dynamics simulation for the transportation means using them, thereby generating flight status data; automatically down-sampling the flight status data to generate a reduced data set; performing high-fidelity flow simulation for the transportation means using the reduced data set; determining in-flight aerodynamic and aero acoustic performance of the transportation means by performing high-fidelity flow simulation; and determining physical characteristics of noise from the transportation means based on the determined in-flight aerodynamic and aero acoustic performance.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] Many existing products and simulation systems are available on the market for designing and simulating objects (e.g., vehicles). Such systems typically use computer-aided design (CAD) and computer-aided engineering (CAE) programs. These systems allow users to build, manipulate, and simulate complex three-dimensional models of objects or assemblies of objects. These CAD and CAE systems provide model representations of objects (e.g., real-world objects) using edges or lines, including faces in certain cases. The lines, edges, faces, or polygons may be represented in various ways, for example, with non-uniform rational B-splines (NURBS). [Background technology]

[0002] These systems manage parts or assemblies of parts of modeled objects, which are primarily specifications of geometric shapes. In particular, a CAD file contains the specifications, from which the geometric shapes are generated. From the geometric shapes, a three-dimensional CAD model or model representation is generated. The specification, geometric shapes, and CAD model / representation may be stored in a single CAD file or multiple CAD files. CAD or other such CAE systems include graphics tools to provide the designer with a visual representation of the modeled object represented in three-dimensional space; these tools are specialized for displaying complex real-world objects. For example, an assembly may contain thousands of parts.

[0003] The advent of CAD and CAE systems has enabled a wide range of representation possibilities for objects, such as CAD models. Computer-based models may be programmed in such a way that they have the properties (e.g., physical, material, or other physics-based) of the real-world object on which they are based or the object they represent. Exemplary properties include stiffness (ratio of force to displacement), plasticity (irreversible strain), and viscosity (resistance to flow of one layer over an adjacent layer), among others. When a CAD, or other such computer-based model known in the art, is programmed in such a manner, it can be used to simulate the object it represents. For example, a mesh-based model can be used to represent the interior cavity of a vehicle, the acoustic fluid surrounding a structure, or any number of real-world objects. Furthermore, CAD and CAE systems, along with computer-based models, can be utilized to simulate engineering systems, such as real-world physical systems (e.g., automobiles, airplanes, buildings, and bridges, among other examples). Furthermore, CAE systems can be used to simulate any of the various behaviors of these physics-based systems, such as noise and vibration, and combinations of the behaviors of these physics-based systems. Summary of the Invention

[0004] Transportation and commuting are fast-growing areas of interest due to continued population growth. A solution to modernizing transportation and alleviating congestion is the use of flying vehicles, such as electric vertical take-off and landing (eVTOL) vehicles. While eVTOLs offer a promising solution to modern transportation problems, little is known about the operation of such vehicles. Embodiments provide a solution to this problem. In particular, embodiments evaluate the noise generated by a vehicle (e.g., an eVTOL) under operating conditions.

[0005] One such embodiment is directed to a computer-implemented method for determining vehicle noise physical characteristics, such as an optimal flight path. According to one embodiment, the method begins by defining a computer-based model of the vehicle and automatically determining the aerodynamic and propulsive performance of the vehicle based on the defined computer-based model. In response, a flight dynamics simulation of the vehicle is performed using the determined aerodynamic and propulsive performance. According to one embodiment, solving the flight dynamics simulation allows tracking of a dynamic equilibrium state at each time step. Running the flight dynamics simulation generates flight situation data, e.g., flight trajectory data. According to one embodiment, during the flight dynamics simulation, a set of flight conditions, e.g., vehicle speed, angle of attack, rotor RPM, etc., that define a "flight situation" are saved to computer disk memory at a given time rate, e.g., every 0.05 seconds of physical flight time. Upon completion of a flight mission, the entire flight situation is considered a "flight envelope." Subsequently, the method automatically downsamples the flight situation data, i.e., the "flight envelope," in an innovative manner, e.g., to generate a reduced data set of flight conditions. A high-fidelity flow simulation (which may be a computational fluid dynamics (CFD) simulation) of the vehicle is then performed using the reduced data set. It is further noted that embodiments may perform multiple such high-fidelity simulations. Performing the high-fidelity flow simulation determines the in-flight aerodynamic and aeroacoustic performance of the vehicle. Thus, the aerodynamic and aeroacoustic performance of the vehicle is determined for a reduced set of representative flight conditions. Subsequently, physical characteristics of the vehicle noise are determined based on the determined in-flight aerodynamic and aeroacoustic performance, and a representation of the determined physical characteristics of the noise is stored in computer memory. According to one embodiment, performing the high-fidelity flow simulation and determining the physical characteristics of the noise are implemented automatically by one or more digital processors.In embodiments, the computer-based model of the vehicle may be defined interactively by a user. Additionally, all steps of the method embodiments may be performed automatically in an automated workflow implemented by one or more processors.

[0006] Another embodiment generates at least one of a propulsion lookup table and an aerodynamic lookup table. According to one embodiment, rotor thrust coefficient and torque coefficient lookup tables are created for different values ​​of rotor RPM and advance rate J, which may be determined from the vehicle geometry. These two tables, rotor thrust coefficient and torque coefficient, form the propulsion performance lookup table and constitute the low-fidelity rotor performance input used by the flight dynamics model. In another embodiment, the aerodynamic performance of the vehicle is determined based on the defined computer-based model by isolating the vehicle airframe from the defined computer-based model, including the aerodynamic control surfaces, and determining aerodynamic performance parameters from the isolated airframe and control surfaces. Aerodynamic force coefficient lookup tables are then created for different flow angle values ​​and different deviation angle values ​​for the various control surfaces. These aerodynamic force coefficient lookup tables form the aerodynamic lookup table and constitute the low-fidelity aerodynamic performance input used by the flight dynamics model.

[0007] One embodiment defines a computer-based model of a vehicle by importing a file containing a digitized representation of the vehicle's geometry. Another embodiment determines propulsive performance of the vehicle based on the defined computer-based model by first isolating one or more rotors of the vehicle from the defined computer-based model and then determining rotor parameters from the isolated one or more rotors. A propulsive lookup table is then searched based on the rotor parameters to determine the propulsive performance. Similarly, one embodiment determines aerodynamic performance of the vehicle based on the defined computer-based model by (1) isolating an airframe of the vehicle from the defined computer-based model, (2) determining aerodynamic parameters from the isolated airframe, and (3) searching an aerodynamic lookup table based on the aerodynamic parameters to determine the aerodynamic performance.

[0008] According to one embodiment, performing a flight dynamics simulation of a vehicle includes defining a travel path. Such an embodiment simulates vehicle operation along the travel path using the determined propulsive and aerodynamic performance, for example, to ensure force balance. Simulating operation, e.g., flight of the vehicle, generates flight situation data at each of a plurality of time steps, and such data is stored in computer memory as flight situation data, i.e., a flight envelope. One embodiment simulates vehicle operation along the travel path using the determined propulsive and aerodynamic performance by providing flight situation data (which may include aerodynamic forces) for a given time step to an autopilot controller. In response, control inputs for the vehicle are received from the autopilot controller, and the received control inputs are used in the simulation to generate flight situation data for time steps following the given time step.

[0009] In one embodiment, the flight situation data includes values ​​for each of a plurality of flight parameters over time. An embodiment downsamples the flight situation data to generate a reduced data set by calculating a multidimensional histogram using the values ​​for each of the plurality of flight parameters over time, where the multidimensional histogram indicates the probability of an in-flight condition. Such an embodiment determines, for each non-zero probability condition, lower and upper bounds for each of the plurality of flight parameters. For each of the plurality of flight parameters, an average lower bound and an average upper bound are calculated based on the determined lower bound and the determined upper bound for each non-zero probability condition. A subset of the calculated average lower and average upper bounds is saved as the reduced data set. Another embodiment determines the subset of the calculated average lower and average upper bounds based on the probabilities of the conditions indicated in the histogram. Yet another embodiment identifies in-flight events based on the histogram and determines the subset of the calculated average lower and average upper bounds based on the identified events.

[0010] According to one embodiment, the high-fidelity flow simulation is a computational fluid dynamics (CFD) simulation or other such fluid simulation known to those skilled in the art. In yet another embodiment, a high-fidelity CFD simulation is performed for all flight / operating conditions within the downsampled data set. In yet another embodiment, performing a high-fidelity flow simulation of the vehicle includes performing multiple high-fidelity flow simulations. In one embodiment, each high-fidelity flow simulation is performed using a respective flight condition from the reduced data set to determine the vehicle's in-flight aerodynamic and aeroacoustic performance for each of the respective flight conditions. One embodiment determines physical characteristics of the vehicle's noise (e.g., in the far field) for each of the respective flight conditions, and selects a given flight condition from among the respective flight conditions based on the physical characteristics of the vehicle's noise determined for each of the respective flight conditions. Another embodiment determines physical characteristics of the noise, such as, for example, the acoustic spectrum and integrated noise level in the far field, as a function of ground distance, atmospheric conditions, and flight context, selected from a set of available flight conditions.

[0011] Another embodiment collects real-world environmental data from one or more sensors. Such an embodiment may perform multiple high-fidelity flow simulations using the respective flight conditions, the reduced data set, and the collected real-world environmental data. Performing the simulations determines the in-flight aerodynamic and aeroacoustic performance of the vehicle subject to the respective flight conditions and the real-world environmental data. Then, for example, the vehicle noise physical characteristics for each of the respective flight conditions over the real-world environment are determined. Next, a given flight condition is selected from among the respective flight conditions based on the vehicle noise physical characteristics determined for each of the respective flight conditions. Further, the selection may be based on noise or other desired performance metrics selected by one or more users. The vehicle is then controlled according to the selected given flight condition.

[0012] One embodiment determines the physical characteristics of the noise of a vehicle at multiple ground locations during flight. Such an embodiment determines the physical characteristics of the noise at the multiple ground locations based on (i) the topography of the ground (i.e., the terrain) at each ground location, (ii) the aerodynamic performance of the vehicle at a waypoint of flight corresponding to each ground location, and (iii) the aeroacoustic performance of the vehicle at a waypoint of flight corresponding to each ground location.

[0013] In one embodiment, one or more high-fidelity flow simulations of the vehicle are performed using the reduced data set. Physical noise metrics are then stored in computer memory for each flight condition determined from the one or more high-fidelity simulations. The stored physical noise metrics are used to calculate noise at one or more ground locations along the flight path. Such an embodiment can also identify an optimal flight mission profile based on ground noise or other performance metrics.

[0014] Another embodiment is directed to a system including a processor and a memory having computer code instructions stored thereon, wherein the processor and memory are configured such that the computer code instructions cause the system to perform any embodiment or combination of embodiments described herein.

[0015] In another embodiment, a computer program product includes a non-transitory computer-readable medium having computer-readable program instructions stored thereon that, when executed by a processor, cause the processor to perform any embodiment or combination of embodiments described herein.

[0016] Yet another embodiment performs a vehicle flight simulation and community noise assessment by first importing a file containing a digitized representation of the three-dimensional vehicle shape. These embodiments separate the rotors (i.e., propellers) from the overall geometry and calculate propulsive performance using a low-fidelity method. These embodiments also separate the airframe from the overall geometry and calculate aerodynamic performance using an automated coarse-mesh CFD simulation. The geometry and mass model are used to set up a flight dynamics model that is used to calculate the vehicle's flight dynamics along a predetermined path. These embodiments sample the flight situation during flight and then downsample the flight envelope based on a histogram to define representative flight conditions for high-fidelity aerodynamic and aeroacoustic CFD calculations. The high-fidelity full vehicle CFD simulation is then performed to calculate aerodynamic and propulsive performance, as well as noise over a microphone hemisphere around the vehicle. These embodiments can repeat the flight simulation using previous or improved aerodynamic and propulsive coefficients and calculate ground noise by far-field extrapolation of the previously calculated noise hemispherical spectrum.

[0017] In one embodiment, importing the file includes calculating the mass and moments of inertia of the vehicle parts using CAD software, and calculating rotor thrust and torque lookup tables by importing a digital representation of the rotor in the form of a stereolithography (STL) file or similar such format. The rotor thrust and torque lookup tables can also be calculated from rotor constructive parameters exported via a CATIA® parametric model of the rotor. In an embodiment, importing the file may also include calculating aerodynamic forces on the airframe by importing a digital representation of the vehicle in the form of an STL file or similar such file.

[0018] One embodiment sets up a flight dynamics model by connecting several modeler modules in a graphical component-based software environment, in one such embodiment, the component-based software environment provides various features including at least one of a world model, an atmosphere model, a terrain model, a vertiport model, a conventional aircraft flight dynamics model, and a multicopter rotor dynamics model, among other examples.

[0019] One embodiment sets up the flight dynamics model by connecting a Dymola® Dynamic Behavior Modeling (DBM) flight dynamics model to a flight controller (e.g., autopilot) and ground control station (GCS) to track and visualize a given flight trajectory. In another embodiment, setting up the flight dynamics model includes looking up a lookup table of force coefficients for the particular airframe and propulsion units of the vehicle being evaluated.

[0020] In another aspect, embodiments downsample the flight envelope by averaging high-speed signals of flight parameters sampled every 0.05 seconds over a larger time window, e.g., 0.5 seconds, calculating histograms of all flight parameters, and ranking conditions accordingly. Embodiments may also identify "corner" events in the flight envelope. Furthermore, embodiments may define the most representative flight conditions for high-fidelity simulation.

[0021] In another aspect, conducting a high-fidelity full vehicle CFD simulation includes automatically generating setup variants for different flight conditions and different control surface settings (flaps, elevators, ailerons, etc.) and tilt rotor angle values. These embodiments extract constructive parameters from the rotor geometry necessary to create partitions on the blades and set up appropriate mesh resolution regions around the rotor. Embodiments also schedule and manage multiple simulations on a high-performance computing (HPC) system, labeling and storing data for each flight condition.

[0022] Embodiments may also calculate noise over a hemisphere (NHD) using an on-the-fly frequency-domain Ffowcs-Williams & Hawkings (FW-H) formulation that accounts for wakes across an integrated surface. Additionally, embodiments may use aerodynamic and propulsive coefficients to train and improve low-fidelity methods through machine learning techniques and heuristic approaches.

[0023] One embodiment performs ground noise calculations offline by importing a complete flight trajectory and corresponding sequence of flight conditions. Another embodiment performs noise calculations on the fly (or in real time) by using the last waypoint of the trajectory and corresponding flight parameters. An embodiment can also interpolate the noise hemisphere from a pre-calculated NND based on instantaneous flight parameters. Still other embodiments apply Doppler effect correction and ground reflection / absorption and atmospheric absorption to the far-field noise calculation at one or more microphone locations on the Earth's surface. An embodiment also determines ground microphone coordinates from a digital representation of the Earth's topography based on the waypoint latitude / longitude coordinates and represents the coordinates as a carpet of microphones of user-defined dimensions that follow the vehicle's ground projection during its flight.

[0024] As part of performing flight mission analysis, one embodiment visualizes the vehicle and noise in realistic rendered scenarios. Another embodiment extends the flight mission analysis framework to eVTOL lifecycle management through the 3DEXPERIENCE® Requirements, Functional, Logical, and Physical (RFLP) capability, a general model-based systems engineering framework that can manage the entire system lifecycle by considering several disciplines and aspects of vehicle design, manufacturing, and maintenance. [Brief explanation of the drawings]

[0025] The foregoing will be apparent from the following more particular description of example embodiments, as illustrated in the accompanying drawings, in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the embodiments.

[0026] [Figure 1] FIG. 1 illustrates a CAD model of an eVTOL vehicle that can be evaluated using embodiments. [Figure 2] FIG. 2 is a flow chart of a method for assessing vehicle noise according to one embodiment. [Figure 3] FIG. 3 illustrates a workflow for performing low-fidelity aerodynamic simulations in one embodiment. [Figure 4] FIG. 4 illustrates the workflow of flight dynamics modeling in one embodiment. [Figure 5] FIG. 5 is a flow diagram illustrating a workflow for flight dynamics modeling with an autopilot in the loop, according to one embodiment. [Figure 6] FIG. 6 illustrates an exemplary user interface that may be employed in an embodiment. [Figure 7] FIG. 7 illustrates a flight envelope downsampling method that may be implemented in an embodiment. [Figure 8]Figure 8 shows the workflow for performing high-fidelity aerodynamic and aeroacoustic simulations. [Figure 9] FIG. 9 illustrates a noise footprint calculation that may be performed in an embodiment. [Figure 10] FIG. 10 is a flow diagram illustrating a method for assessing vehicle noise according to one embodiment. [Figure 11] FIG. 11 is a simplified block diagram of a computer system embodiment for determining vehicle noise. [Figure 12] FIG. 12 is a simplified diagram of a computer network environment in which embodiments of the present invention may be implemented. DETAILED DESCRIPTION OF THE INVENTION

[0027] A description of an exemplary embodiment follows.

[0028] Embodiments evaluate the noise generated by a vehicle, for example, an electric vertical take-off and landing (eVTOL) vehicle, in an operating scenario. It is noted that while embodiments are described herein with respect to eVTOLs, the embodiments are not limited to evaluating the noise of eVTOLs and may be used to evaluate the noise of any vehicle flight.

[0029] As urban populations continue to grow, traffic is becoming increasingly problematic. Millions of hours are wasted on roads worldwide, resulting in significant personal and business losses. The combined use of air taxi vehicles (ATVs), personal aerial vehicles (PAVs), and small package delivery drones has the potential to radically improve commuter mobility and positively impact society, quality of life, and the environment. Therefore, the development of urban air mobility (UAM) concepts, from vehicle design to traffic management, has attracted the interest of several stakeholders, including aircraft and automotive manufacturers, regulatory authorities and agencies, research institutes, and academic institutions, among others. The rapid growth of this topic, in terms of industrial investment, research and development efforts, and media coverage, is typically referred to as the third aerospace revolution.

[0030] However, before such technologies can be widely adopted, challenges must be overcome in the areas of: (i) vehicles (e.g., battery durability and propulsion efficiency, noise and safety), (ii) manufacturing (e.g., serial production of composite structures to aerospace quality standards), (iii) operation (e.g., specific air traffic management (ATM) systems capable of managing both manned / unmanned and guided / autonomous vehicles), (iv) regulation (e.g., related to certification, operation, and safety), (v) life cycle management and maintenance, and (vi) recycling.

[0031] Considering the interplay between different disciplines involved in the design of safe and quiet flight systems, such as eVTOLs, it can be argued that a new design paradigm is needed at all levels, from conceptual to detailed full-model design. Similar to helicopters rather than fixed-wing aircraft, the design of different systems is tightly coupled to the aerodynamic behavior of the vehicle. More specifically, due to the strong link between maneuvers and noise perceived on the ground, and between installed power and trajectory, the dynamic trim of the vehicle is a key element of any design and optimization process. Embodiments predict the impact of trajectory and maneuvers on the noise generated by the vehicle. Embodiments can be implemented as part of a flight mission platform and can be operated in real time to control and influence the real-world operation of eVTOLs or such flight vehicles.

[0032] Optimizing the takeoff and landing procedures of eVTOL vehicles is a challenging and interesting task due to the large number of flight and operational control parameters compared to helicopters. The target mission profile being considered by multiple stakeholders is a 60-mile air shuttle between an airport and a central business district. An analysis of battery capacity revealed that lifting surfaces are required to limit the generation of electric lift during takeoff and landing and to cover the target range. Therefore, the most viable eVTOL concept is a vehicle lifted by rotors that translate from a vertical axis to a horizontal axis thanks to tilting rotors and / or tilting lifting surfaces.

[0033] 1 shows an example of an eVTOL concept vehicle 100 whose noise can be evaluated using embodiments described herein. The exemplary vehicle 100 comprises two forward counter-rotating open rotors 101a-b and 101c-d and two aft shrouded counter-rotating tilted rotors 102a-b. The ability to tilt rotor and wing components constitutes an additional degree of freedom in the aerodynamic and aeroacoustic optimization of the maneuver, making it a challenging geometry to work with for simulation and noise evaluation purposes.

[0034] In a multi-rotor configuration, blade-vortex interaction (BVI) phenomena can occur under certain flight conditions when the wake from one rotor is captured by another rotor. Thus, for this type of architecture (vehicle 100), low noise procedures are those that minimize the occurrence of BVI conditions.

[0035] Electric motors allow the use of variable-speed rotors, which allow the rotor's rotational speed to be adapted to different flight conditions to provide the required lift and thrust, while operating close to the optimum advance rate, i.e., the ratio between tip speed and freestream fluid velocity. Compared to controlling the collective angle of helicopter rotors, rotational speed control can be used to better distribute lift between the wing and rotor during the transition phase, i.e., from hover to cruise conditions.

[0036] Based on these aforementioned factors, an eVTOL flight from point A to point B can be accomplished in many different ways, with some trajectories and operating parameters being noisier than others. A goal of embodiments is to assess the ground noise footprint of an eVTOL during its flight by considering both aerodynamic and aerodynamic effects.

[0037] 2 is a flowchart of an exemplary method 220 for one such embodiment for assessing vehicle noise. Method 220 begins, at 221, by defining a computer-based model of the vehicle and, at 222, automatically determining the aerodynamic and propulsive performance of the vehicle based on the defined computer-based model. In response, a flight dynamics simulation of the vehicle is performed, at 223, using the determined aerodynamic and propulsive performance determined in step 222. Performing flight dynamics simulation 223 generates flight situation data. This flight situation data may constitute a flight envelope, i.e., equilibrium states along a given trajectory or flight.

[0038] This flight situation data is automatically downsampled in step 224 to generate a reduced data set. A high-fidelity flow simulation of the vehicle is then performed in step 225 using the reduced data set, i.e., the flight conditions represented in the reduced data set output from step 224. Performing the high-fidelity flow simulation 225 determines the in-flight aerodynamic and aeroacoustic performance of the vehicle. In one embodiment, multiple high-fidelity simulations are performed in step 225 such that the in-flight aerodynamic and aeroacoustic performance of the vehicle for every set of flight conditions in the reduced data set is determined.

[0039] The vehicle noise physical characteristics are then determined based on the determined in-flight aerodynamic and aeroacoustic performance, and a representation of the determined noise physical characteristics is stored in computer memory in step 226. In one embodiment, for a given flight, ground noise is calculated in step 226.

[0040] As noted, method 220 is computer-implemented, such that functions and operations, e.g., steps 221-226, may be implemented automatically by one or more digital processors. Furthermore, method 220 may be implemented using any computer device or combination of computing devices known in the art. Among other examples, method 220 may be implemented using computer system 1100, described later in this specification in connection with FIG. 11, and computer network environment 1200, described later in this specification in connection with FIG. 12.

[0041] One embodiment of method 220 defines a computer-based model of the vehicle in step 221 by importing a file containing a digitized representation of the vehicle's geometry. This digitized representation may be in the form of a CAD model or a discrete surface, such as a tessellated geometry (stl). Additionally, step 221 of method 220 may implement the functionality of step 1, described later in this specification in connection with FIG. 10.

[0042] Embodiments of method 220 may generate at least one of a propulsion lookup table and an aerodynamic lookup table. One such embodiment creates rotor thrust and torque coefficient lookup tables for different values ​​of rotor RPM and advance rate J, which may be determined from the vehicle geometry represented by the computer-based model defined in step 221. These two tables, rotor thrust and torque coefficient, form the propulsion lookup table and constitute the low-fidelity rotor performance inputs used by the flight dynamics model. In another embodiment, the aerodynamic performance of the vehicle is determined based on the defined computer-based model 221 by isolating the vehicle airframe from the defined computer-based model, including the aerodynamic control surfaces, and determining aerodynamic parameters from the isolated airframe and control surfaces. Aerodynamic force coefficient lookup tables are then created for different flow angle values ​​and different deviation angle values ​​for the various control surfaces. These aerodynamic force coefficient lookup tables form the aerodynamic lookup tables and constitute the low-fidelity aerodynamic performance inputs used by the flight dynamics model.

[0043] According to an embodiment of method 220, determining the propulsive performance of the vehicle (222) based on the defined computer-based model includes first isolating (i.e., identifying, extracting, etc.) one or more rotors of the vehicle from the defined computer-based model. Second, rotor parameters are determined from the isolated one or more rotors, for example, by a segmentation process and analysis, after which an aerodynamic panel method is applied. Then, in step 222, a propulsive lookup table is searched based on the rotor parameters to determine the propulsive performance. Similarly, embodiments may determine the aerodynamic performance of the vehicle in step 222 by isolating the vehicle's airframe from the defined computer-based model in step 221. Such embodiments determine aerodynamic parameters from the isolated airframe and search an aerodynamic lookup table based on the aerodynamic parameters to determine the aerodynamic performance in step 222.

[0044] Additionally, embodiments may, at step 222, implement workflow 330, or portions thereof, described later herein in conjunction with Figure 3, to determine the aerodynamic and propulsive performance of the vehicle. Additionally, step 222 of method 220 may implement the functionality of steps 2 and 3, described later herein in conjunction with Figure 10, to determine the aerodynamic and propulsive performance of the vehicle.

[0045] One embodiment of method 220 performs a flight dynamics simulation of the vehicle in step 223 by first defining a travel path, such as a theoretical line from A to B. Such an embodiment uses the propulsive and aerodynamic performance determined in step 222 to ensure equilibrium and simulates the vehicle's motion along the travel path, such as by solving one or more equations of motion. This simulation 223 generates flight situation data for the vehicle at each of a plurality of time steps. The flight situation data at each of the plurality of time steps is stored as flight situation data, i.e., a flight envelope for the particular vehicle.

[0046] One embodiment of method 220 simulates vehicle operation along a travel path using the determined propulsive and aerodynamic performance characteristics by providing flight situation data for a given time step to an autopilot controller in step 223. In response, control inputs for the vehicle are received from the autopilot controller. The received control inputs are used in the simulation to generate flight situation data for time steps following the given time step. To illustrate, consider an example in which the simulation is performed one time step at a time in step 223. In such an embodiment, at a given time step, e.g., 1-1.5 seconds, the vehicle has the same set of specific characteristics as the environment in which the vehicle is operating. These characteristics, i.e., flight situation data, are provided to the autopilot controller, which determines that a characteristic, e.g., rotor speed, needs to be modified based on the data and the desired flight path or other such specifications. This modification is provided by the autopilot controller to software implementing the simulation (223), and the vehicle characteristics for the next time step, 1.5-2 seconds, are updated accordingly, i.e., rotor speed is increased.

[0047] An embodiment of method 220 may, at step 223, implement workflows 330, 440, and 550 (or portions thereof) described later herein in conjunction with Figures 3, 4, and 5, respectively, to perform a flight dynamics simulation to generate flight situation data. Additionally, at step 223, an embodiment may implement the functionality of step 4 of workflow 1000 described later herein in conjunction with Figure 10 to perform a flight dynamics simulation.

[0048] In one embodiment, the flight situation data generated by simulation 223 includes values ​​for each of a plurality of flight parameters over time, i.e., a discretized flight path. One embodiment of method 220 generates a reduced dataset by downsampling the flight situation data in step 224 to calculate a multidimensional histogram over time using values ​​for each of the plurality of flight parameters. In such an embodiment, the multidimensional histogram indicates the probability of a condition during flight. One embodiment determines, for each non-zero probability condition, lower and upper bounds for each of the plurality of flight parameters. For each of the plurality of flight parameters, an average lower bound and an average upper bound are calculated based on the determined lower bound and the determined upper bound for each non-zero probability condition. A subset of the calculated average lower and average upper bounds is saved as a reduced dataset. According to one embodiment, the output dataset, i.e., the reduced dataset, is typically several factors smaller than the input dataset, i.e., the flight situation dataset.

[0049] To illustrate, consider a simplified four-parameter, non-limiting example in which the parameters are rotor speed, rotor tilt angle, flight Mach number, and flight angle of attack. At the start of flight, the rotor speed (RPM) is likely at a maximum value to generate takeoff lift, the tilt angle is 90 degrees (vertical axis), the Mach number is zero, and the flight angle of attack is −90 degrees (flowing from above). The vehicle remains in this state for the entire vertical climb, which means that this is the state that will be considered for further noise calculations and should be narrowed down for further high-fidelity flow (e.g., CFD) simulations. Upon reaching a given altitude, the vehicle begins transition to horizontal flight. During this phase, the Mach number gradually increases and the rotor speed likely gradually decreases. These transient flight conditions during the transition to horizontal flight can generate a large number of parameter combinations as the engine nacelles are progressively tilted from a vertical to a horizontal setting and the angle of attack varies from −90 degrees to values ​​close to the nominal cruise value. A given number of flight conditions, e.g., combinations of these parameters, are narrowed down to a selection range for further high-fidelity CFD simulations. At cruise conditions, nominal settings are maintained by the controller, and these nominal settings correspond to one (or a few) conditions that will be considered for further high-fidelity CFD simulations. During turns, other parameters, such as control surface deflections, come into play, but the concept remains similar. Finally, during the approach phase, which involves the transition from horizontal to vertical flight, a situation similar to the climb phase detailed above occurs.

[0050] Another embodiment of method 220 determines the subset of calculated average lower and upper bounds based on the probability of the conditions indicated by the histogram. For example, such an embodiment may consider only conditions that exceed a certain probability threshold. Yet another embodiment identifies in-flight events based on the histogram and by a priori listing exceptions based on vehicle type, and determines the subset of calculated average lower and upper bounds based on the identified events. For example, such an embodiment may ensure that data for such identified events are included in the subset by a priori listing them according to vehicle type. Furthermore, an embodiment of method 220 implements the functionality of workflow 770 described later in this specification in connection with FIG. 7 in step 224 to downsample the data and create a reduced data set. In another embodiment, the functionality of step 5 of workflow 1000 described later in this specification in connection with FIG. 10 is implemented in step 224 to downsample the flight situation data.

[0051] According to one aspect of method 220, the high-fidelity flow simulation performed in step 225 is a computational fluid dynamics simulation. In such an embodiment, the computational fluid dynamics simulation is performed in step 225 using a reduced data set according to principles known to those skilled in the art. One embodiment of method 220 implements the functionality of workflow 880, described later herein in connection with FIG. 8, in step 225 to perform the high-fidelity flow simulation. In another embodiment, the functionality of step 6 of workflow 1000, described later herein in connection with FIG. 10, is implemented in step 225 to perform the high-fidelity flow simulation.

[0052] Another embodiment performs testing in step 225 to determine optimized flight control parameters and / or conditions. Such an embodiment performs multiple high-fidelity flow simulations in step 225, each using a respective flight condition from the reduced data set. These respective flight conditions are different parameters to be tested. Performing the multiple high-fidelity flow simulations determines the vehicle's in-flight aerodynamic and aeroacoustic performance for each of the respective flight conditions. Such an embodiment can then identify a set of optimized respective flight conditions for operating the vehicle in the real world. For example, one embodiment of method 220 determines the vehicle's noise physical characteristics for each of the respective flight conditions and selects a given flight condition from among the respective flight conditions based on the determined vehicle's noise physical characteristics for each of the respective flight conditions. For example, a flight condition that minimizes noise may be selected.

[0053] Another embodiment of method 220 collects real-world environmental data from one or more sensors. Such an embodiment performs multiple high-fidelity flow simulations using the respective flight conditions, the reduced data set, and the collected real-world environmental data to determine the in-flight aerodynamic and aeroacoustic performance of the vehicle subject to the respective flight conditions and the real-world environmental data. Physical characteristics of the vehicle noise for each of the respective flight conditions are then determined. A given flight condition is then selected from among the respective flight conditions based on the determined physical characteristics of the vehicle noise for each of the respective flight conditions. One embodiment then controls the real-world vehicle according to the selected given flight condition.

[0054] Yet another embodiment of method 220 collects real-world environmental data from one or more sensors and uses this data to perform a new flight simulation in step 226. Flight conditions along the flight path are used to identify and extract corresponding noise characteristics, which are calculated in step 225 and stored in computer memory. From the extracted physical noise metrics, noise radiated at one or more locations on the ground is calculated in step 226. By considering different flight paths or parts thereof (specific maneuvers), it is possible to identify optimal behavior based on noise or other performance metrics.

[0055] In one embodiment of method 220, determining the physical characteristics of the vehicle noise in step 226 includes determining the physical characteristics of the vehicle noise at multiple ground locations during the flight. Such an embodiment determines the physical characteristics of the noise at the multiple ground locations based on (i) ground topography at each location, (ii) the aerodynamic performance of the vehicle at a waypoint of the flight corresponding to each ground location (i.e., the aerodynamic performance of the vehicle when in an airborne position substantially directly above the point on the ground), and (iii) the aeroacoustic performance of the vehicle at a waypoint of the flight corresponding to each ground location (i.e., the aeroacoustic performance of the vehicle when in an airborne position substantially above the point on the ground).

[0056] Further, in step 226, an embodiment of method 220 implements the functionality of workflow 990 described later in this specification in conjunction with FIG. 9 to determine the physical characteristics of the noise. According to one embodiment, the noise is determined based on the determined in-flight aerodynamic and aeroacoustic performance by utilizing a noise footprint calculation using a hemispherical approach. Determining the physical characteristics of the noise in step 226 may include determining the noise radiated at one or more points on the ground. According to one embodiment, the process of determining the noise radiated at one or more points on the ground includes tracing a ray connecting the instantaneous position of the vehicle, the center of the hemisphere, and the point on the ground, and interpolating the noise from the NHD at the intersection between the ray and the hemisphere. If ground reflection is applied, secondary rays can be considered. The noise level interpolated from the NHD is then corrected by applying range broadening correction, Doppler correction, atmospheric absorption, and ground reflection. In another embodiment, the functionality of step 7 of workflow 1000 described later in this specification in conjunction with FIG. 10 is implemented in step 226 to determine the noise characteristics. According to one embodiment, the physical characteristics of the noise include noise power spectral density levels (units: dB / Hz) in discrete narrow frequency bands at several points on a hemisphere around the vehicle. In one embodiment, for all flight conditions, i.e., combinations of flight parameters, the hemisphere is stored in computer memory and tagged with the specific condition.

[0057] Low-Fidelity Aerodynamics

[0058] An embodiment can employ a low-fidelity aerodynamic simulation workflow 330, as illustrated in FIG. 3. It is noted that for simplicity, FIG. 3 only shows in-plane forces and angles of attack. In one embodiment, the low-fidelity simulation 330 performs simulations using an informed blade element momentum theory (BEMT) model to calculate rotor performance. In one embodiment, the BEMT model used in workflow 330 is based on standard theory with several parameters adjusted based on high-fidelity computational fluid dynamics (CFD) calculations, as described later in this specification. These simulations are used to populate the rotor lookup table 335. Another embodiment of the low-fidelity simulation workflow 330 uses a Multicopter Aerodynamic and Aeroacoustic Simulation (MAAS) automated process to generate simulation setups for PowerFLOW®, the high-fidelity CFD simulation software by applicant-assignee Dassault Systemes Simulia Corporation. The automation automatically controls the resolution of the computational mesh in the optimal range to generate sufficiently accurate predictions with a reasonable computational turnaround time. The low-fidelity attribute of workflow 330 does not refer to simplifications to the physical model or geometry, but rather to the use of a coarser mesh resolution. The results of these simulations may be used to populate lookup table 338. In this manner, embodiments perform simulations and store the results used in flight dynamics simulations. This storage reduces processing time during future implementations.

[0059] During operation, the low-fidelity aerodynamic simulation 330 controls the propulsion system geometry 339a, the airframe geometry 339b, and the rotor RPM 339c, the rotor advance rate J 339d, and the flight Mach number M ∞The system begins with input data 339, which includes flight / operation parameter ranges for flow incidence angle (angle of attack) α 339e, and flow incidence angle (angle of attack) α 339f. Geometry information 339a-b and ranges for the aforementioned parameters 339c-f are then provided to the simulation software described above, or such other simulation software, to determine output 340. In one embodiment, if a simulation using the software has previously been performed, no additional simulation is performed to determine output 340; instead, lookup tables 335 and 338 are searched to determine output 340, taking into account input data 339. In one embodiment, output 340 is determined by a thrust coefficient C T 340a, torque coefficient C Q 340b, lift coefficient C L 340c, drag coefficient C D 340d, and moment coefficient C M Includes 340e.

[0060] This low-fidelity simulation 330 uses the geometries of the main vehicle components, namely rotor / propeller geometry 331 and airframe geometry 332, to calculate coefficients of performance over a wide range of operating conditions, for example, using the coarse mesh simulation described above. One embodiment stores the propulsion parameters of rotor 331 in lookup table 335, and calculates the thrust coefficient C, respectively. T 340a and torque coefficient C Q 340b, which depends on the revolutions per minute (RPM) and the advance rate J between the axial flow velocity (flight speed) and the root number n times the rotor diameter D (J=V / n / D). Lookup table 335 considers different values ​​of RPM and J and indicates the resulting thrust coefficient 340a and torque coefficient 340b, respectively. In one embodiment, these values ​​are stored in a lookup table, such as the illustrated rotor lookup table 335, for subsequent use.

[0061] The aerodynamic parameters of the airframe 332 are the flight Mach number M ∞, the lift coefficient C along three major axes (not shown), which depends on the angle of attack a, and the crosswind angle β. L 340c and drag coefficient C D 340d, side force coefficient (not shown), and moment coefficient. For simplicity, FIG. 3 only shows lift 340c, drag 340d, and pitch 340e moment coefficients as a function of angle of attack. In one embodiment, lookup table 338 calculates the lift 340c, drag 340d, and pitch 340e moment coefficients as a function of angle of attack. ∞ , and crosswind angle β.

[0062] Thus, workflow 330 performs a simulation, e.g., a CFD simulation, using a coarse mesh having characteristics described by input data 339. The resulting output data 340 is stored in lookup tables 335 and 338. In one embodiment, if suitable data exists or can be determined, for example, through interpolation, no simulation is performed; instead, lookup tables 335 and 338 are searched based on input data 339 to determine the resulting output 340.

[0063] Flight Dynamics Modeling

[0064] 4 illustrates a flight dynamics modeling workflow 440 according to one embodiment. Workflow 440 uses the Dassault Systemes Dymola® behavior modeling (DBM) process to solve flight dynamics equations to determine non-inertial equilibrium for all points in the trajectory connecting a starting point, e.g., a vertiport (point A), to a target destination (point B). Dymola® has the ability to integrate Modelica flight dynamics and rotor performance models with other Modelica models (world, atmosphere, terrain, vertiport) that describe operational scenarios.

[0065] In workflow 440, interface 441 is used to provide input data for the simulated flight by solving flight dynamics balance equations. This input data 442 includes vehicle part weights 442a, target trajectory 442b, world model 442c, atmosphere model 442d, terrain model 442e, vertiport model 442f, flight mechanics model 442g, multicopter rotor dynamics model 442h, rotor thrust and torque table 442i, and airframe lift and drag table 442j. This input data 442 is used by the Dymola® platform to determine flight conditions along trajectory 443.

[0066] In another embodiment of workflow 440, the vehicle part weights and moments of inertia required to solve the flight dynamics equations are received directly from the Dassault Systemes CATIA® computer-aided design component based on the original vehicle model directly linked to a Dynamic Behavior Modeling (DBM) library.

[0067] 5 illustrates another workflow 550 that may be used for flight dynamics modeling according to one embodiment. Workflow 550 connects workflow 440 model (442b) with an autopilot computer, e.g., a third-party autopilot computer, that tracks a target trajectory by acting on control parameters (e.g., vehicle control surfaces and rotor RPM) and receives flight situations from the DBM model in a closed-loop process according to a conventional software-in-the-loop (SIL) process.

[0068] In an exemplary implementation of workflow 550, at 551, the motion parameters, e.g., output 443, determined using workflow 440, are provided to an autopilot calculator, which reads the parameters at 552. At step 553, the autopilot calculator compares the motion parameters to the desired kinematics to determine new actuator positions (e.g., corrections to the vehicle), which are sent (at 554) to the Dymola® platform, which updates the vehicle characteristics (e.g., input data 440) according to the new actuator positions at 555, uses the updated data to determine the motion parameters, and returns to step 551. In this manner, the avionics of the virtual model are used in a feedback loop with the DBM model to model the flight dynamics of the vehicle.

[0069] In one embodiment of workflow 550, the autopilot computer interacts with a virtual ground control station (GCS) as well as a virtual graphical cockpit model, and these graphical components are coupled with the DBM and Dassault Systemes 3D scenario visualization tools (e.g., Creative Experience applications) within the 3DEXPERIENCE® flight mission simulator, as illustrated in Figure 6. Figure 6 shows the resulting interface 660, which acts as an output visualization and includes, for example, virtual cockpit plots 661a-d, a GCS visualization 662, and a flight simulator representation 663 for rotor speed, rotor tilt angle, attitude angle, and airspeed, respectively.

[0070] Flight envelope downsampling

[0071] 7 illustrates a workflow 770 for downsampling data. Workflow 770 begins with receiving flight situation data (input 771) for a flight mission. In one example, input data 771 includes values ​​for Mach number 771a, angle of attack 771b, crosswind angle 771c, rotor RPM 771d, rotor tilt angle 771e, and control surface angle 771f, sampled every 0.05 seconds. This input data 771 represents a flight envelope, and workflow 770 downsamples the flight envelope data 771 to generate a reduced number of combinations 772 of input data 771. After downsampling 770, a reduced number of conditions 772 are identified under which a high-fidelity simulation of the vehicle's aerodynamic behavior is performed. In one embodiment, downsampling 770 involves initial filtering and averaging of the raw data, followed by identification of the most likely flight conditions and "corner" events based on vehicle type and prior knowledge, and finally a graphical check of the original flight envelope cover. Plots 773 and 774 illustrate the filtering, where in plots 773 and 774 the original flight situation data is shown as dots and the downsampled flight situation data is shown as crosses.

[0072] High-fidelity aerodynamics and aeroacoustics

[0073] 8 illustrates a workflow 880 for conducting a high-fidelity aerodynamic and aeroacoustic simulation of a vehicle 884. Workflow 880 begins with input data 881, which, according to one embodiment, includes propulsion system geometry 881a, airframe geometry 881b, and downsampled flight conditions 881c. Input data 881a-c is used to conduct a high-fidelity simulation that generates output data 882 for all flight conditions from downsampled data 881c. Exemplary output data 882 includes rotor thrust and torque coefficients 882a, airframe aerodynamic coefficients 882b, and noise hemisphere data 882c.

[0074] In one embodiment, workflow 880 includes scheduling and running multiple high-fidelity aerodynamic and aeroacoustic calculations on a high-performance computing (HPC) cloud system using PowerFLOW® (a Lattice Boltzmann-based transient CFD flow calculation) and OptydB_FOOTPRINT (a far-field noise calculation by applicant-assignee Dassault Systemes Simulia Corporation). In this embodiment, a scheduler is part of the MAAS workflow implementing the embodiment, and the scheduler manages and performs data labeling and storage. In such an embodiment, for all flight conditions (from downsampled data 881c), MAAS automatically creates a PowerFLOW® simulation setup, submits the job on the HPC system, and performs data analysis. The primary output of the aerodynamic calculations is the forces acting on portions of the vehicle, which are time-averaged and used to train a low-fidelity model used as part of the low-fidelity simulations described herein. Other outputs include flow pressure and density, as well as velocity variations across multiple surfaces around the entire vehicle. In an exemplary embodiment, this data is used by the tool "optydb_FWHFREQ" by applicant-assignee Dassault Systemes Simulia Corporation to calculate the noise over a hemisphere of microphones distributed around the vehicle and fixed to the vehicle reference system. The tool optydb_FWHFREQ can manage a large number of microphones (on the order of 1000) and remove the spurious effects of the vehicle's wake across the integrated fluid surface. The optydb_FWHFREQ tool can provide these features by employing a specific Ffowcs-Williams & Hawkings (FW-H) formulation that includes a quadrupole noise correction in the frequency domain.This feature, together with the average of the noise spectra calculated from different layers of the synthesis surface (typically three layers as shown in FIG. 8 by the cylinder 883 surrounding the vehicle simulation setup), allows workflow 880 to determine a very accurate noise spectrum over the hemisphere. In one embodiment, the spectra of all microphones over the hemisphere for all flight conditions are exported to a file, all together constituting a noise hemisphere database from which ground noise can be extrapolated.

[0075] Noise footprint calculation

[0076] Figure 9 illustrates a noise footprint calculation workflow 990 according to one embodiment. Workflow 990 begins with input data 991, which includes input data 442 / 991a from flight dynamics workflow 440 described herein above in connection with Figure 4, and a noise hemisphere database 991b, e.g., output data 882c from workflow 880. This input data 991 is used to generate output noise footprint data 992. The output is a visualization 993a-993b showing the total sound pressure level (in dB) integrated over a user-specified frequency range. b According to one embodiment, the visualization 993a- b is created using noise hemisphere database technology 994. In one embodiment, output 992 includes instantaneous noise levels on the ground 992a, cumulative noise metrics 992b, and audio files 992c for the waypoints flown.

[0077] According to one embodiment, workflow 990 performs flight mission simulation 440, described herein above in connection with FIG. 4, while sending flight conditions to tool optydb_FOOTPRINT every 0.5 seconds (or at some other desired frequency). In such an implementation, tool optydb_FOOTPRINT interpolates the corresponding noise hemispherical spectrum based on the three closest flight conditions contained in NHD 991b. The interpolated noise hemispherical spectrum is then extrapolated to the Earth's surface by applying Doppler correction, atmospheric absorption, and ground reflection / absorption to obtain the ground noise level shown in visualizations 993a-b in FIG. 9. In an embodiment, the noise calculation can be performed offline by importing the flight trajectory once the flight is completed, or the noise calculation can be performed in real time / on the fly by using the last updated vehicle position. The rate of on-the-fly noise calculation can be increased to match real-time stepping as needed.

[0078] In one embodiment, workflow 990 advantageously extracts ground points at the start of the calculation or on the fly by considering the current vehicle position, for example, for short-term event flights such as vehicle landings. In one embodiment, ground points are extracted from the Shuttle Radar Topography Mission (SRTM) Earth topography database using a proprietary Dassault Systemes tool that takes as input the coordinates of the vehicle where the noise carpet is centered, as well as the X and Y dimensions of the carpet. When the on-the-fly noise calculation is performed, the carpet follows the vehicle and the noise levels can be visualized on the 3DEXPERIENCE® platform.

[0079] The functionality described herein above in connection with Figures 3-9 are elements that may be employed in embodiments (e.g., method 220) to assess vehicle noise. Figure 10 illustrates a workflow 1000 that combines the functionality described herein above in connection with Figures 3-9 to provide eVTOL flight mission analysis and community noise forecasting. Workflow 1000 includes the following steps: Step 1. Create a vehicle geometry and set up the Dymola® aerodynamic model (shape, weight, moments of inertia, etc.). Step 2. Use the low-fidelity optydb_BEMT approach to calculate the rotor thrust and torque coefficients to obtain C T and C Q Get the lookup table. Step 3. Calculate the airframe aerodynamic coefficients using the coarse PowerFLOW® setup automated by MAAS and compare the calculated coefficients with C L , C Q , C Y etc., stored in a lookup table. Step 4. Conduct a Dymola®-based flight simulation to collect the flight envelope along the entire mission. Step 5. Downsample the flight envelope using a statistical approach to obtain a reduced flight envelope. Step 6. Run PowerFLOW® simulations for reduced flight envelope flight conditions and calculate vehicle ambient noise spectra using optydb_FWHFREQ. Stored in the global Noise Hemisphere Database (NHD). Step 7. Repeat the flight mission simulation on the same or a similar trajectory and use optydb_FOOTPRINT to calculate the ground noise on the fly. Step 8. A machine learning approach uses the high-fidelity results to train and improve the low-fidelity results.

[0080] With reference to FIG. 10, the following describes the different stages of a flight mission analysis and community noise prediction workflow 1000 according to one embodiment.

[0081] Workflow 1000 - Step 1

[0082] Workflow 1000 begins in step 1 with the preparation of the eVTOL shape of vehicle 1001 using the CATIA® suite of computer-aided three-dimensional interactive applications from Dassault Systemes. One embodiment uses specific modules of CATIA® to evaluate the mass moments of inertia of parts of vehicle 1001, which are required by the CATIA® DBM model to solve the flight dynamics equations along the flight. In an exemplary implementation, CATIA® and Dymola® are connected by the 3DEXPERIENCE® / CATIA® / Functional & Logical Design application by Dassault Systemes. It is noted that while specific software applications and platforms are described herein as being employed by embodiments, embodiments are not limited to such applications and platforms, and embodiments may be implemented using any of a variety of known software applications and platforms.

[0083] In one embodiment, the Dymola® DBM module solves systems of flight dynamics differential algebraic equations (DAEs). More specifically, Dymola® relies on "component-based modeling" supported by a graphical user interface (GUI), where systems are described by blocks and connecting lines that represent "acausal" dependencies between components. This approach allows for the easy construction of components and subsystems, similar to how engineers build real systems. A component-based modeling language allows for the modeling and integration of physical systems across multiple domains. An object-oriented paradigm is used by Dymola® to create acausal connections between components, thereby enabling the flow of information in all directions. The object-oriented language employed by Dymola® is Modelica, an equation-based modeling language, whereby models are declared through equations from first principles, without the need to specify how to calculate them. Modelica is also a non-proprietary language and has a large repository of open-source libraries across a variety of disciplines. For the same reasons, there are numerous Modelica simulation environments available, both commercially and freely. One of the main Modelica modules employed in an embodiment of workflow 1000 is the FlightDynamics library, developed by the German Aerospace Center (DLR). This library contains basic components for building various types of aircraft models, as well as environment models for simulating aircraft under operating conditions. The Modelica Standard Library, including the MultiBody library, is used in conjunction with the FlightDynamics library to model eVTOL dynamic behavior.

[0084] A useful feature of Modelica for modeling dynamic systems such as eVTOLs is its ability to automatically convert nonlinear direct models into nonlinear inverse models. The inverse model of the eVTOL system is used for trim calculations and automatic generation of control laws, even for flexible bodies such as wings. The trimming phase of flight simulation allows the simulation to determine command states, such as angular deflections of working surfaces or rotor RPM, which ensures satisfaction of dynamic equilibrium during flight.

[0085] In one embodiment, other Modelica modules are integrated into the eVTOL flight mission analysis framework to describe operational scenarios (world, atmosphere, terrain, vertiports), and the modules communicate using the so-called Functional Mockup Interface (FMI), a standard for co-simulation. The modules are also linked to two external (third-party) components: the autopilot, which generates commands to be applied to flight surfaces and other control parameters based on the difference between the vehicle's actual position and the target reference trajectory. The commands are used by the DBM to determine the vehicle's new dynamic state (forces and moments) and new position after a discrete, small time interval (e.g., 0.05 seconds). Another third-party component is the Ground Control Station (GCS), which defines the trajectory and facilitates visualization of the vehicle's state and position along its mission.

[0086] To seamlessly utilize multiple tools according to the framework described herein, a common platform on which these tools are available is advantageous. In the exemplary implementation of workflow 1000, the 3DEXPERIENCE® platform serves this purpose. In one embodiment, the tools are provided by the platform in the form of roles consisting of applications. Each of these applications has specific functionality for a different part of the project, and some of the applications that share a certain synergy can be used together. In the exemplary implementation of workflow 1000, three such applications are used to perform the process: (1) a DBM application; (2) a Functional & Logical design application used to interface the DBM model with the flight controller, GCS, and 3D visualization software; and (3) a Creative Experience application used for real-time 3D visualization of the simulation.

[0087] The functional and logical applications are part of the 3DEXPERIENCE® Requirements, Functional, Logical, and Physical (RFLP) framework, which is based on the V-cycle design process. Therefore, the proposed framework can be extended to the systems engineering level by considering the following aspects: (i) requirements: establishing criteria for verification, validation, and qualification of the vehicle design; (ii) functional: defining the services or technical capabilities that the vehicle will provide to meet the requirements; (iii) logical: defining the logical architecture of the eVTOL using its components, their relationships, and behaviors; and (iv) physical: representing the virtual solution of the eVTOL.

[0088] In the proposed framework, logical nodes are used for the purpose of interfacing different components. However, the use of functional and logical design applications allows the proposed framework to be expanded to cover the complete product development cycle of eVTOLs.

[0089] The DBM application can update the vehicle position based on the vehicle's current kinematic state (speed, flow angle) and the command values ​​provided by the flight controller, as in a typical SIL approach (as shown in workflow 550). The commands are used to determine the variation of forces acting on a component, such as the thrust generated by a rotor at a given RPM and advance rate. In workflow 1000, the aerodynamic forces acting on the airframe and generated by the rotor are provided to the DBM via pre-calculated lookup tables. These lookup tables cover a variety of possible operating conditions of the vehicle in flight, so in the first case, these tables are created using a low-fidelity method.

[0090] Workflow 1000 - Step 2

[0091] For rotor thrust and torque, the Dassault Systemes tool optydb_BEMT is used in step 2. The tool optydb_BEMT is based on a validated blade element momentum theory with lift and drag coefficients of the blade cross section calculated using an incorporated proprietary module. This tool optydb_BEMT calculates C, which is a function of rotor advance rate J and RPM. T (J,RMP)1003 and C Q (J,RPM) 1002. Exemplary embodiments use either a digital representation of the rotor, such as a triangular stereolithography file (STL), or constructive parameters exported by a parametric CATIA® model of the blades.

[0092] Furthermore, it is noted that in embodiments of workflow 1000, if the required data is available in tables 1002 and 1003 or if the required data can be derived from tables 1002 and 1003, then there is no need to perform a simulation in step 2.

[0093] Workflow 1000 - Step 3

[0094] For the aerodynamic forces acting on the airframe, a low-fidelity model is provided by the Multicopter Aerodynamic and Aeroacoustic Simulation (MAAS) workflow by Dassault Systemes, which is used to automatically set up a complete eVTOL system simulation via the high-fidelity CFD software SIMULIA / PowerFLOW®. The attribute of low fidelity relates to PowerFLOW®'s ability to automatically generate a computational mesh with a resolution value such that simulation costs covering the entire flight envelope are affordable. This approach allows for the use of effective vehicle geometry rather than surrogate models of components. Furthermore, thanks to the parameterization of the setup creation, it is possible to simulate different conditions by changing the numbers in the MAAS input file. These different conditions can be used in step 3 to generate aerodynamic force lookup tables 1004, 1005, 1006, and 1007. Furthermore, it is noted that in embodiments of workflow 1000, if the required data is available in tables 1004-1007 or the required data can be derived from tables 1004-1007, then there is no need to perform a simulation in step 3.

[0095] Workflow 1000 - Step 4

[0096] Once the aerodynamic force lookup tables 1002-1007 are available and the DBM model is available and linked to the flight controller and GCS, it is possible to prescribe a trajectory and simulate the flight. The controller can integrate constraints, but at this stage, the primary concern is flying and collecting vehicle conditions along the flight. This data is collected by sampling and recording control parameters (e.g., rotor RPM, tilt rotor angle, working surface angle, etc.) and flight conditions (e.g., Mach number, angle of attack, side wing angle, etc.) every 0.05 seconds (or at some other desired frequency) to create a dataset. This dataset can be as large as 10,000 samples and is highly dependent on the duration and complexity of the flight path. This dataset constitutes a subset of the flight envelope for a given flight mission and may include data for different phases, including vertical takeoff, vertical-to-horizontal transitions, main flight with latitude changes and several turns, horizontal-to-vertical transitions, and vertical landing. Visualizations 1010 and 1011 show the visual output of step 4.

[0097] Workflow 1000 - Step 5

[0098] Due to the wide variety of conditions, the dataset collected in step 4 is quite large. Workflow 1000 addresses the size of the dataset in step 5 by downsampling the flight envelope dataset and defining sample conditions (e.g., flight path data, i.e., trajectory 1024 with vehicle points 1025, e.g., points at which noise is calculated), as visually illustrated in step 5. Embodiments may perform downsampling based on a variety of different considerations. For example, embodiments may identify sample conditions, and therefore noise conditions (corner events), based on condition specificity based on incidence (histogram) or load.

[0099] These two criteria, occurrence rate and condition specificity, allow for the definition of a predetermined number of flight conditions to be used in high-fidelity aerodynamic and aeroacoustic simulations. Prior to constructing the histogram, one embodiment forces the angle of attack value to zero whenever the Mach number falls below a certain threshold and averages high-frequency samples over a large time window of 0.5 seconds. This allows for filtering out unusual events that are not representative of average dynamic behavior and improves the robustness of the downsampling process. This filtering is illustrated in plots 1021 and 1022, where plot 1021 shows the original data and plot 1022 shows the filtered data.

[0100] According to one embodiment, the histogram-based process for downsampling in step 5 of workflow 1000 includes the steps of: (i) dividing the value range of each parameter based on a step value provided by the user; (ii) calculating the multidimensional histogram and the probability of all interval combinations; (iii) searching for non-zero probability conditions and saving their indexes; (iv) looping through the non-zero probability conditions and collecting the lower and upper parameter values ​​for these conditions; and (v) calculating the average values ​​of those lower and upper limits and saving them to an output file. Embodiments can also consider corner events of flight, regardless of their probability, to better define the boundaries of the flight envelope.

[0101] Workflow 1000 - Step 6

[0102] In step 6, a high-fidelity simulation is performed using the downsampled dataset generated in step 5. Depending on the available computational resources, a given number of flight conditions (parameter combinations) taken from the ranked subset (downsampled dataset) are considered. The considered flight conditions are used to automatically set up a high-fidelity PowerFLOW® simulation using MAAS, which includes both the airframe and rotors of the vehicle 1020 operating at a specific RPM value. The high-fidelity simulation in step 6 is performed by scheduling multiple jobs in an HPC system and using a cloud-based GUI to monitor the job status and label and store the simulation results. In one embodiment, a high level of automation in MAAS is achieved through the use of the optydb_PFROTOR tool by applicant-assignee Dassault Systemes Simulia Corporation, which facilitates the extraction of constructive parameters from the rotor geometry required to generate an appropriate computational mesh around the rotor.

[0103] PowerFLOW® transient flow calculations provide unsteady aerodynamic forces on different parts of the vehicle, as well as flow variables (pressure, density, and velocity components) on multi-layered sampling surfaces around the vehicle. One large file of several hundred gigabytes is generated per run, containing all time steps. The physical time covered by each simulation is typically 10-20 rotor revolutions. The sampling rate is on the order of 1000 BPF / B, where BPF is the blade passing frequency, e.g., the rotational frequency multiplied by the number of rotor blades, B. The noise generated by the vehicle is calculated on the fly (not in post-processing as is typically done) using the frequency-domain FW-H tool optydb_FWHFREQ.

[0104] Using this approach, embodiments can calculate noise spectra from as many as 1000 microphones on a hemisphere around a vehicle in a much more efficient manner than using a time-domain approach. To save storage space, users can decide to enable an option to automatically delete FW-H flow input files at the end of a simulation. In addition to on-the-fly capabilities, the FW-H model can also reduce the spurious effects of vortex wakes that cross integration surfaces. This is achieved using two simultaneous techniques. The first technique uses multiple adjacent integration surfaces (typically three) to average the complex acoustic spectrum. The second technique uses a formula that takes into account the contributions of otherwise missing volume sources through an additional surface integral that approximates the missing volume integral.

[0105] The noise spectrum for each microphone M on the hemisphere is stored as a narrowband power spectral density value PSD(f,M). The entire set of all files for the downsampled flight envelope data constitutes the Noise Hemisphere Database (NHD) that is used for the noise footprint calculation in step 7 of the workflow.

[0106] In one embodiment, the unsteady aerodynamic force signals exported by each simulation run are averaged in time to obtain steady-state aerodynamic and thrust / torque rotor coefficients on the airframe. These values ​​are used to construct a correction metamodel that is applied to the low-fidelity model. According to one embodiment, two different approaches are pursued for the rotor and airframe. For the rotor, a machine learning-based algorithm is used to modify tuning parameters of the BEMT model, e.g., the topic correction coefficients and the 3D correction coefficients for the cross-sectional lift and drag coefficients. For the airframe components, the main error is related to the absence of rotating components in the low-fidelity calculation; therefore, the effects of rotor-induced flow are evaluated by subtraction logic and used to feed a heuristic correction model. The corrected low-fidelity model can be used in step 8 of the workflow.

[0107] Workflow 1000 - Step 7

[0108] Step 7 of workflow 1000 involves repeating the flight mission using the same DBM model as step 6. However, an improved lookup table for aerodynamic forces is used in step 7. Optionally, different trajectories may be defined, provided that they correspond to similar flight envelopes previously calculated. In one embodiment, the flight controller may be replaced by a human pilot in a flight simulation Human-In-the-Loop (HIL) modality. Furthermore, step 7 may consider two different scenarios for noise calculation: (i) focusing on a fixed region on the ground that is fixed during the flight event; and (ii) focusing on a moving region on the ground that follows the vehicle ground projection during flight.

[0109] In both cases, a noise carpet, i.e., a grid of ground microphones, is extracted from a discrete representation of the Earth from the Shuttle Radar Topography Mission (SRTM) database based on latitude / longitude coordinates. This is accomplished through a tool that exports the coordinates as a cloud of points or a mosaic surface in Universal Transverse Mercator (UTM) format, readable by the noise footprint calculation tool optydb_FOOTPRINT. In the first case (a fixed area on the ground), carpet extraction is performed only once at the beginning of the flight mission analysis, while in the second case (focusing on moving areas), extraction is performed at discrete time intervals. Another difference between these two analysis modalities is that in the first case, time-accumulated noise metrics such as sound exposure level (SEL) and effective perceived noise level (EPNL) can be calculated at the end of the flight event. This approach is particularly useful for optimizing specific phases of a flight, such as takeoff and landing.

[0110] Noise calculations may be performed after flight simulation, using trajectory waypoints and operating conditions (parameter combinations) at every waypoint, or during flight using the last waypoint and corresponding operating conditions. Depending on the available computing power and the time step used for noise calculations, which may be much larger than the actual flight simulation time step, flight simulation and noise calculations may be performed in real time, in either SIL or HIL modalities. Step 7 may also produce a visualization 1026 in which contours, i.e., shading, indicate different ground noise levels below the vehicle 1027. In one embodiment, visualization 1026 is generated using noise calculation techniques 1028.

[0111] In one embodiment, the optydb_FOOTPRINT tool performs the following operations to calculate the noise at the ground microphones: First, the NHD is imported and saved in memory. Second, for each combination of flight parameters corresponding to a waypoint, the three closest conditions stored in the NHD are determined via a suitable multidimensional distance. Third, an equivalent noise hemisphere is interpolated based on the NHD. Fourth, the noise map calculation is distributed among multiple cores. Fifth, the hemisphere is rotated to take into account the vehicle's instantaneous Euler angles. Sixth, the intersection between the rotated hemisphere and a ray connecting the vehicle's delayed position (at the release time) and a point on the ground is calculated. Seventh, the noise at the intersection is interpolated based on the hemispherical microphones. Eighth, the noise PSD from the hemisphere to the ground is extrapolated by applying corrections to account for atmospheric absorption depending on distance, frequency, air humidity, Doppler amplitude and frequency corrections (which depend on the Mach number projected along the radiation direction), and ground absorption and reflection (which depend on the sound-absorbing properties of the terrain). Ninth, a noise visualization metric is calculated, for example, the overall sound pressure level (OSPL) at all microphones on the ground. Tenth, a noise map is constructed that can be projected onto the terrain and visualized within Creative Experience applications. The ground noise PSD for each discrete time is also saved to a file, which can be used for offline noise metric calculations such as SEL or EPNL.

[0112] Workflow 1000 - Step 8

[0113] In step 8 of workflow 1000, the high-fidelity results are used to train and improve the low-fidelity model used in steps 2 and 3. In one embodiment, machine learning methods are used to implement this training so that better lookup tables can be created in the future.

[0114] Embodiments provide a computational method and system for conducting a community noise assessment of a vehicle under realistic operating conditions. The embodiments are based on a flight mission analysis that solves the aerodynamic dynamic equilibrium of the vehicle along its flight. A software-in-the-loop strategy with an automatic controller can be used to track a user-specified flight path. The aerodynamic forces generated by the flow over the airframe and the thrust / torque generated by each rotor at RPM values ​​set by the controller are estimated by an aerodynamic solver from lookup tables. To cover a wide range of operating conditions, the lookup tables are calculated using a low-fidelity method. During the simulation flight, the entire set of flight parameters is recorded at high speed (e.g., 0.05 seconds). After the first flight, a statistical argument-based procedure is used to reduce the number of flight conditions to a representative, user-specified set of conditions. These conditions are then used to conduct high-fidelity, full-vehicle aerodynamic and aeroacoustic calculations using an automated process for pre-processing, execution, and post-processing of the simulation. The high-fidelity noise spectrum over the hemisphere around the vehicle calculated during the aerodynamic simulation is stored in a database for further ground noise calculations. The high-fidelity aerodynamic results are used to correct the low-fidelity lookup table for further flight simulations. Finally, a new flight mission, which may differ from the first one, is simulated, and the instantaneous flight conditions are used to calculate the ground noise from the pre-calculated noise database.

[0115] Acoustic calculations can be performed in post-processing after the flight is completed or during the flight (on-the-fly analysis). The rate of the acoustic calculations can be adjusted so that ground noise can be calculated and visualized in real time. Noise maps can be calculated on fixed ground patches extracted from an Earth topography database at the start of the flight simulation, or on ground patches that follow the vehicle along its flight and can be extracted from the Earth topography database for each time step of the noise calculation.

[0116] Computer Support

[0117] FIG. 11 is a simplified block diagram of a computer-based system 1100 that can be used to implement any of the various embodiments of the invention described herein. System 1100 includes a bus 1103. Bus 1103 serves as an interconnect between the various components of system 1100. Connected to bus 1103 is an input / output device interface 1106 for connecting various input and output devices, such as a keyboard, mouse, display, and speakers, to system 1100. A central processing unit (CPU) 1102 is connected to bus 1103 and provides for the execution of computer instructions to implement embodiments. Memory 1105 provides volatile storage for data used in the execution of computer instructions to implement embodiments described herein, such as those previously described hereinabove. Storage 1104 provides non-volatile storage for software instructions, such as an operating system (not shown) and the configuration of embodiments. System 1100 also includes a network interface 1101 for connecting to any of the various networks known in the art, including wide area networks (WANs) and local area networks (LANs).

[0118] It should be understood that the exemplary embodiments described herein may be implemented in many different ways. In some cases, the various methods and systems described herein may each be implemented by a physical, virtual, or hybrid general-purpose computer, such as computer system 1100, or a computer network environment, such as computer environment 1200 described later in this specification in connection with FIG. 12. Computer system 1100 may be converted into a system that performs the methods described herein (e.g., 220, 330, 440, 550, 770, 880, 990, 1000), for example, by loading software instructions into either memory 1105 or non-volatile storage 1104 for execution by CPU 1102. Those skilled in the art should further understand that system 1100 and its various components may be configured to perform any embodiment or combination of embodiments of the invention described herein. Furthermore, system 1100 may implement the various embodiments described herein utilizing any combination of hardware, software, and firmware modules operably coupled internally or externally to system 1100.

[0119] 12 illustrates a computer network environment 1200 in which an embodiment of the present invention may be implemented. In the computer network environment 1200, a server 1201 is linked to clients 1203a-n via a communications network 1202. The environment 1200 may be used to enable the clients 1203a-n, alone or in combination with the server 1201, to perform any of the embodiments described herein. In a non-limiting example, the computer network environment 1200 provides a cloud computing embodiment, a software as a service (SAAS) embodiment, or the like.

[0120] The embodiments or aspects thereof may be implemented in the form of hardware, firmware, or software. If implemented in software, the software may be stored on any non-transitory computer-readable medium configured to allow a processor to load the software, or a subset of its instructions. The processor then executes the instructions and is configured to operate or cause an apparatus to operate in the manner described herein.

[0121] Furthermore, firmware, software, routines, or instructions may be described herein as performing certain operations and / or functions of a data processor, although it will be understood that such description contained herein is merely for convenience and that in reality such operations result from a computing device, processor, controller, or other device executing the firmware, software, routines, instructions, etc.

[0122] It should be understood that the flow diagrams, block diagrams, and network diagrams may include more or fewer elements, or may be arranged differently, or may be represented differently, but it should also be understood that the particular implementation may dictate the number of block diagrams and network diagrams that illustrate the execution of an embodiment implemented in a particular way.

[0123] Accordingly, further embodiments may also be implemented in a variety of computer architectures, physical, virtual, cloud computers, and / or any combination thereof, and therefore the data processors described herein are intended for illustrative purposes only and not as limitations of the embodiments.

[0124] The teachings of all patents, published applications, and references cited herein are incorporated by reference in their entirety.

[0125] While exemplary embodiments have been particularly shown and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the embodiments encompassed by the appended claims.

[0126] For example, the foregoing description and details of the illustrated embodiments refer to applicant-assignee (Dassault Systemes Simulia Corporation) and Dassault Systemes tools and platforms for purposes of illustration, but not limitation. Other similar tools and platforms are suitable.

Claims

1. 1. A computer-implemented method for determining physical characteristics of vehicle noise, the method comprising: defining a computer-based model of the vehicle; automatically determining aerodynamic and propulsive performance of the vehicle based on the defined computer-based model; and responsively performing a flight dynamics simulation of the vehicle using the determined aerodynamic and propulsive performance, wherein performing the flight dynamics simulation generates flight situation data including values ​​for each of a plurality of flight parameters over time; calculating a multidimensional histogram using the values ​​for each of the plurality of flight parameters over time, the multidimensional histogram indicating a probability of a condition during flight; determining, for each non-zero probability condition, lower and upper bounds for each of the plurality of flight parameters; calculating an average lower limit value and an average upper limit value for each of the plurality of flight parameters based on the determined lower limit value and the determined upper limit value for each non-zero probability condition; storing a subset of the calculated average lower bound and average upper bound values ​​as a reduced data set; automatically downsampling the flight situation data to generate a reduced data set by: performing a high-fidelity flow simulation of the vehicle using the reduced data set, the high-fidelity flow simulation determining the in-flight aerodynamic and aeroacoustic performance of the vehicle; determining noise physical characteristics of the vehicle based on the determined in-flight aerodynamic and aeroacoustic performance, and storing a representation of the determined noise physical characteristics in computer memory, wherein performing the high-fidelity flow simulation and determining the noise physical characteristics are performed automatically by one or more digital processors.

2. defining the computer-based model of the vehicle; The method of claim 1 , comprising importing a file containing a digitized representation of the vehicle geometry.

3. determining aerodynamic and propulsive performance of the vehicle based on the defined computer-based model; Isolating one or more rotors of the vehicle from the defined computer-based model; determining rotor parameters from the isolated one or more rotors; calculating a propulsion lookup table based on the rotor parameters to determine the propulsion performance; Isolating the vehicle airframe from the defined computer-based model; and determining aerodynamic parameters from the separated airframe; and calculating an aerodynamic look-up table based on the aerodynamic parameters to determine the aerodynamic performance.

4. The method of claim 3 , further comprising generating and storing at least one of the propulsion lookup table and the aerodynamic lookup table in computer memory.

5. conducting the flight dynamics simulation of the vehicle; Defining a travel path; simulating movement of the vehicle along the travel path using the determined propulsive and aerodynamic performances, wherein simulating the movement generates flight situation data at each of a plurality of time steps; and storing the flight situation data at each of the plurality of time steps as the flight situation data.

6. simulating operation of the vehicle along the travel path using the determined propulsive and aerodynamic performances; providing flight situation data for a given time step to an autopilot controller; responsively receiving control inputs for the vehicle from the autopilot controller; and generating flight situation data for a time step subsequent to the given time step using the received control input in the simulating.

7. The method of claim 1 , further comprising determining the subset of the calculated average lower and upper bound values ​​based on probabilities of conditions indicated by the histogram.

8. identifying the in-flight event based on the histogram; and The method of claim 7 , further comprising: determining the subset of the calculated average lower and upper bound values ​​based on the identified events.

9. The method of claim 1 , wherein the high-fidelity flow simulation is a computational fluid dynamics (CFD) simulation.

10. performing the high fidelity flow simulation of the vehicle; 10. The method of claim 1, comprising performing a plurality of high-fidelity flow simulations, each high-fidelity flow simulation being performed using a respective flight condition from the reduced data set to determine in-flight aerodynamic and aeroacoustic performance of the vehicle for each of the respective flight conditions.

11. determining a noise physical characteristic of the vehicle for each of the respective flight conditions; 11. The method of claim 10, further comprising: selecting a given flight condition from among the respective flight conditions based on the determined noise physical characteristics of the vehicle for each of the respective flight conditions.

12. collecting real-world environmental data from one or more sensors; performing a plurality of high-fidelity flow simulations, each high-fidelity flow simulation being performed using a respective flight condition, the reduced data set, and the collected real-world environmental data to determine in-flight aerodynamic and aeroacoustic performance of the vehicle subject to the respective flight condition and the real-world environmental data; determining a noise physical characteristic of the vehicle for each of the respective flight conditions; selecting a given flight condition from among the respective flight conditions based on the determined noise physical characteristics of the vehicle for each of the respective flight conditions; The method of claim 1 , further comprising: controlling the vehicle according to the selected given flight conditions.

13. determining a physical characteristic of the vehicle noise; 10. The method of claim 1, comprising determining physical characteristics of noise of the vehicle at a plurality of ground locations during flight, wherein the physical characteristics of noise at the plurality of ground locations are determined based on (i) ground topography at each ground location, (ii) aerodynamic performance of the vehicle at waypoints of the flight corresponding to each ground location, and (iii) aeroacoustic performance of the vehicle at waypoints of the flight corresponding to each ground location.

14. 1. A computer system for determining physical characteristics of vehicle noise, said system comprising: a processor; and a memory having computer code instructions stored therein, wherein the processor and the memory use the computer code instructions to cause the system to: defining a computer-based model of the vehicle; automatically determining aerodynamic and propulsive performance of the vehicle based on the defined computer-based model; and responsively performing a flight dynamics simulation of the vehicle using the determined aerodynamic and propulsive performance, wherein performing the flight dynamics simulation generates flight situation data including values ​​for each of a plurality of flight parameters over time; calculating a multidimensional histogram using the values ​​for each of the plurality of flight parameters over time, the multidimensional histogram indicating a probability of a condition during flight; determining, for each non-zero probability condition, lower and upper bounds for each of the plurality of flight parameters; calculating an average lower limit value and an average upper limit value for each of the plurality of flight parameters based on the determined lower limit value and the determined upper limit value for each non-zero probability condition; storing a subset of the calculated average lower bound and average upper bound values ​​as a reduced data set; automatically downsampling the flight situation data to generate a reduced data set by: performing a high-fidelity flow simulation of the vehicle using the reduced data set, the high-fidelity flow simulation determining the in-flight aerodynamic and aeroacoustic performance of the vehicle; determining a noise physical characteristic of the vehicle based on the determined in-flight aerodynamic and aeroacoustic performance, and storing a representation of the determined noise physical characteristic in computer memory.

15. In performing the flight dynamics simulation of the vehicle, the processor and the memory use the computer code instructions to cause the system to: Defining a travel path; simulating movement of the vehicle along the travel path using the determined propulsive and aerodynamic performances, wherein simulating the movement generates flight situation data at each of a plurality of time steps; and storing the flight situation data at each of the plurality of time steps as the flight situation data.

16. In determining the physical characteristics of the vehicle noise, the processor and the memory are configured by the computer code instructions to cause the system to:

15. The system of claim 14, configured to determine physical characteristics of noise of the vehicle at multiple ground locations during flight, wherein the physical characteristics of noise at the multiple ground locations are determined based on (i) ground topography at each ground location, (ii) aerodynamic performance of the vehicle at waypoints of the flight corresponding to each ground location, and (iii) aeroacoustic performance of the vehicle at waypoints of the flight corresponding to each ground location.

17. 1. A computer program product for determining physical characteristics of vehicle noise, said computer program product comprising: one or more non-transitory computer-readable storage devices; and program instructions stored in at least one of the one or more storage devices, the program instructions, when loaded and executed by a processor, causing a device associated with the processor to: defining a computer-based model of the vehicle; automatically determining aerodynamic and propulsive performance of the vehicle based on the defined computer-based model; and responsively performing a flight dynamics simulation of the vehicle using the determined aerodynamic and propulsive performance, wherein performing the flight dynamics simulation generates flight situation data including values ​​for each of a plurality of flight parameters over time; calculating a multidimensional histogram using the values ​​for each of the plurality of flight parameters over time, the multidimensional histogram indicating a probability of a condition during flight; determining, for each non-zero probability condition, lower and upper bounds for each of the plurality of flight parameters; calculating an average lower limit value and an average upper limit value for each of the plurality of flight parameters based on the determined lower limit value and the determined upper limit value for each non-zero probability condition; storing a subset of the calculated average lower bound and average upper bound values ​​as a reduced data set; automatically downsampling the flight situation data to generate a reduced data set by: performing a high-fidelity flow simulation of the vehicle using the reduced data set, the high-fidelity flow simulation determining the in-flight aerodynamic and aeroacoustic performance of the vehicle; determining a noise physical characteristic of the vehicle based on the determined in-flight aerodynamic and aeroacoustic performance, and storing in computer memory a representation of the determined noise physical characteristic.

Images