|Next-Generation NATO Reference Mobility Model (NRMM) Development
|Applied Vehicle Technology|
Ground Vehicle, Mobility, Next-Generation, NRMM, Physics-Based
The NATO Reference Mobility Model (NRMM) is a simulation tool aimed at predicting the capability of a vehicle to move over specified terrain conditions. NRMM can be used for on-road and cross-country scenarios, it can account for several parameters such as terrain type, moisture content, terrain roughness, vehicle geometry, driver capabilities, etc.
NRMM was developed and validated by the U.S. Army Tank Automotive Research, Development, and Engineering Center (TARDEC) and Engineer Research and Development Center (ERDC) over several decades, and has been revised and updated throughout the years, resulting in the most recent version, NRMM II (arising partly out of AVT-107). NRMM is traditionally used to facilitate comparison between vehicle design candidates and to assess the mobility of existing vehicles under specific scenarios.
Although NRMM has proven to be of great practical utility to the NATO forces, when compared to modern modeling tools it exhibits several inherent limitations:
- It is based on empirical observations, and therefore extrapolation outside of test conditions is difficult or impossible.
- It is heavily dependent on in-situ soil measurements.
- Only one-dimensional analysis is possible; lateral vehicle dynamics are not considered.
- It does not account for vehicle dynamic effects, but instead only considers steady-state condition.
- It is specific to wheeled/tracked vehicles.
- It is not easily implementable within modern vehicle dynamics simulations.
- It exhibits poor (or poorly understood) inter-operability and inter-scalability with other terramechanics and soil mechanics models.
- It is only suitable for mobility analysis, and does not provide auxiliary outputs (e.g. power efficiency analysis).
The proposed capability development is vital to NATOs mission. It promises to enable new capabilities in the design, modeling, and simulation of a broad class of vehicles. These modeling capabilities are of high importance to current and future NATO missions because they have the potential to significantly reduce costs and improve performance. The new tool will be applicable to various running gear morphologies, including conventional wheels and tracks, and more novel bio-inspired limb designs. This could yield a new paradigm for ground vehicle mobility, which surpasses traditional analysis based on NRMMs go/no-go basis. An important aspect of modern simulations is the possibility to model complex vehicle maneuvering in high fidelity. Relying on High Performance Computing (HPC), it will be possible to utilize statistical representations of terrain profile and properties and to exploit very large-scale Monte Carlo simulations to yield rich outputs over a broad parameter space.
The scope is to investigate an efficient simulation-based Next-Generation NRMM. Specifically, the proposed activity will focus on the following fundamental scientific objectives:
-Identify scale-invariant terrain descriptions for representing topographic map data (obtained at various scales) within a suitable multi-body dynamic simulator. This will enable automated analysis of regions of interest, given heterogeneous map data products as inputs.
-Develop efficient, automated, parallelizable experimental design methods (i.e. sampling methods) for extracting metrics of interest from Monte Carlo simulations of the multi-body dynamic simulator, including mobility-related metrics and auxiliary metrics. This will yield rich statistical mobility-related outputs in a computationally efficient manner, which will allow use of modern HPC resources.
-Explore the use of compact representations of vehicle dynamics (i.e. response surface methods or other approximation methods) within the multi-body dynamic simulator, with a goal of further reducing computational cost.
-Establish compact, user-friendly representations of output metrics that capture important dependencies. This will yield an update to classical speed made good or go/no-go maps.
Modernizing the NRMM involves several areas of effort:
-Identification of vehicle - terrain interaction models, i.e., terramechanics models, that balance fidelity with computational efficiency. These models may be semi-empirical, or fully analytical based on discrete elements.
-Development of in-situ and online measurement tools to identify required terrain parameters.
-Identification of the type and form of desired responses, to yield rich mobility predictions and (ideally) useful auxiliary outputs.
-Integration of terramechanics models into modern dynamic simulation software, and develop efficient, automated computation tools, which will ideally enable the use of high performance computation techniques.
-Since the Next-Generation NRMM is expected to be extremely computationally intensive, there exists a need to investigate numerical methods to improve algorithmic efficiency and automate NRMM output generation, such as Monte Carlo sampling techniques and stochastic response surfaces.