AIAA Applied Aerodynamics Technical Committee
AIAA Applied Surrogate Modeling Workshop (AASM)
Missile
Missile Case​​​
​
​Task: Predict Roll, Pitch and Yaw rates for the Army Research Lab Laboratory Technology Vehicle (LTV) - 1.​​
​
This database provides aerodynamic forces and moments across a trade space of Mach number, roll angle, angle of attack, as well as deflection angle of the control surfaces, serving as a benchmark for mission engineering, design optimization, and uncertainty quantification. Demonstrated on a missile, the database includes forces and moments as well as the roll, pitch, and yaw rates to various aerodynamic conditions.
​
​
​
​
​
​
​
The underlying benchmark vehicle is the Army Research Lab's Laboratory Technology Vehicle (LTV-1), which designed for flight over a wide envelope from high subsonic to hypersonic speeds. The vehicle features slender body, with four long strakes that enhance lift and four trailing edge flaps at the tail in line with each strake.
​
​
With an initial dataset of 200 samples, the missile database is continuously expanding, providing a scalable and adaptable resource for engineers and researchers developing next-generation surrogate modeling capabilities.
Laboratory Technology Vehicle
This example is intended to provide multi-fidelity data for the integrated aerodynamic loading coefficients that could be used for a 3DOF model of the Laboratory Technology Vehicle (LTV) missile. The US Army Combat Capabilities Development Command (DEVCOM) Army Research Laboratory (ARL) introduced the LTV as an engineering testbed [1]. The vehicle geometry, shown below, was developed for public release, making it an accessible benchmark case for testing.
​
The dataset includes:
· 1000 low-fidelity data points are generated with Missile DATCOM using Latin Hypercube sampling.
· 200 high-fidelity data points are generated with Kestrel using Monte-Carlo Voronoi (MCV) sampling, which is a sampling technique that seeks to enhance the quality of the sampling set. See reference [2] for more details of the MCV implementation.
· 7x11 grid (77 data points) of high-fidelity samples generated as a holdout test data set, intended to assess the model accuracy.
The distribution of these three sample sets is shown below.
​
The datasets are available in standard comma-separated-value (*.csv) format or HDF5 format (*.hdf5). A sample reader for the hdf5 data is included:
1. LTV_LF_1000lhs – 1000 low fidelity Missile DATCOM calculations with Latin Hypercube Sampling
2. LTV_HF_200mcv – 200 high fidelity Kestrel calculations with Monte-Carlo Voronoi sampling
3. LTV_HF_77grid – 7x11 (77pts) high fidelity Kestrel calculations over grid, intended as holdout data for testing.
4. read-hdf5.py – python code to read HDF5 file formats
​
​
References
[1] Vasile, J., J. Bryson, J. Sahu, J. L. Paul, B.C. Gruenwald, “Aerodynamic Dataset Generation of a Long-Range Projectile,” Army Research Lab Publication ARL-TR-9019, Aug. 2020.
[2] Kaminsky, A.L, A. House, L. Jensen, M. Ilu, W. Chapman, A. Brown, A. Wissink, N. Hariharan, D. McDaniel, “Design and Application of the Sage Surrogate Modeling Software,” AIAA SciTech 2025 Forum, Orlando, FL, Jan 6-10, 2025.
​
​​
​
​
​
​
​
​
-
Andrew Wissink, Matthew Liu, Alec House, Andrew Kaminsky, Andrew Simin, and Rakesh Ranjan, "Multi-Fidelity Aerodynamics Surrogate Modeling Using HPCMP CREATE Sage," AIAA SCITECH 2025 Forum, DOI: 10.2514/6.2025-0039​
-
B.T. Burchett, J.L. Paul, J.D. Vasile, J.T. Tyson, “A High Fidelity Roll Dependent Model for a Long- Range High-Speed Missile”, ARL-TR-9236, JULY 2021, https://apps.dtic.mil/sti/trecms/pdf/AD1141614.pdf
Reference
​
Please refer to these papers.​