Keywords

1 Background

Anthropometric casualty assessment consists of methodologies and algorithms that: (1) use of accurate three-dimensional (3D) models of soldiers and PPE; (2) determine interaction of munitions with PPE, taking into account fit and form; and (3) determine casualties from munitions effects. The design of PPE for force protection is critical for soldier survivability and effectiveness for a range of combat operations. The Army Science and Technology (S&T) strategic direction includes the Soldier and Small Unit (SU) Force Protection, Technology Enabled Capability Demonstration (TECD), which identifies the critical need for new technologies to increase protective gear performance by reducing PPE weight and volume, providing protection from weapon threats and ultimately reducing the number and severity of soldier injuries and casualties (Association of United States Army 2011). Protective equipment is subject to many design considerations; the most fundamental are “form” and “fit.” Form is the PPE design elements that are derived by functional choices, such as which areas of the body must be protected for anticipated combat operations. In determining the optimal form, tradeoffs must be made – for example, between level of protection and limits on weight. Fit is how well a specific individual is protected by the form and how the individual’s freedom of motion is impeded by the form. To achieve the designed level of protection, proper fit is essential. Optimal fit maximizes protection while minimizing encumbrance. Less-than-optimal fit can result in inadequate protection or obstruct the soldier’s ability to perform operational tasks.

Lt. Col. Jon Rickey, product manager for soldier protective equipment at Program Executive Office (PEO) Soldier, stated in a 2011 interview with National Defense magazine that having accurate anthropometric sizes of soldiers to ensure proper fit is the crucial first step in designing future body armor (Beidel 2011). The Army has invested significant time and resources into collecting accurate 3D data of active-duty soldiers. Through the most recent activity, the U.S. Army Anthropometric Survey II (ANSUR II), the Army has collected over 12,000 3D scans as of May 2012. Anthropometrists collect this data using laser scanning to produce high-resolution 3D models of each soldier. This technology also can capture high-resolution scans of PPE body armor such as vests or helmets in specified sizes.

With new Army sizing charts under development as part of ANSUR II, the Army needs advanced analysis tools to evaluate the impact of current and future PPE sizing decisions. For example, the Interceptor Body Armor (IBA) Improved Outer Tactical Vest (IOTV) comes in 11 sizes, with modular configurations for lower back, groin, and deltoid protective pieces based on the soldier’s mission. The Advanced Combat Helmet (ACH) comes in five sizes and adds a protective pad between the bottom of the helmet shell and the top of the IBA collar. PPE designers need sophisticated analysis tools to improve form and fit metrics to improve soldier protection.

Live-fire testing of all permutations of PPE sizes, configurations, and individual body types is impractical, but computer Modeling and Simulation (M&S) provides a valid, safe way to study armor effectiveness in preventing or reducing injury from combat hazards such as fragmenting explosives. To analyze PPE fit for a soldier population, it is insufficient to place a static equipment model over a static human model. Soldiers in the battlefield frequently are in motion, and changes in posture may cause the area of PPE coverage to change. The soldier’s position and posture at the moment of an attack ultimately determines if the armor is in the right place to stop an incoming bullet or fragment. To account for variations in PPE coverage that may result from less-than-optimal fit, the equipment model must predict variations in fit due to shifts in posture when applied to digital soldiers.

Fit and PPE effectiveness are influenced greatly by the dynamics of the combat situation. Anthropometric casualty assessment should provide a capability to examine fit in operational scenarios in which soldier agents are engaged in a combat mission. For example, the Infantry Warrior Simulation (IWARS) is a force-on-force constructive combat simulation, because it is the current state-of-the-art for evaluating equipment performance in operational scenarios. Developing anthropometric casualty estimation in a simulation like IWARS will have clear benefits for the evaluation and procurement of future equipment systems.

Furthermore, linking ANSUR II anthropometric data, improved digital models of humans and equipment, and state-of-the-art computer simulations will provide scientists and engineers engaged in the design and analysis of current and future PPE with powerful new tools to assess the protective capabilities of alternative armor configurations. Testing variations in PPE size and configuration in simulation against a representative soldier population enables scientists to use the simulation results to identify potential gaps in armor coverage caused by inadequate fit.

2 The Need for Anthropometric Casualty Estimation

We conducted comparative analysis of existing tools to identify the current state of the art, enabling technologies, and potential transition partners. We examined the Integrated Casualty Estimation Methodology (ICEM) (Simulation Technologies, Inc. 2003) platform and its use of the ORCA (Neades et al. 1998) casualty estimation model. Based on this research, we determined that the body models being used by ICEM had no capability to account for individual body variations. Additionally, we consulted with potential transition partners to determine needs and capability gaps in current analysis tools. The Army identified a need to be able to perform analysis with a constructive simulation that will enable analysts to substitute their own models for certain components, including alternative casualty assessment models and munitions flyout models.

3 Analysis Process to Study PPE Fit and Form

We developed an analysis process consisting of three primary analytical components: (1) input; (2) DHM/DEM configuration; and (3) output and analysis components. This analysis process is illustrated in (Fig. 1).

Fig. 1.
figure 1

The ACES analysis process

Full size image

The development of the analysis process led to the construction of a system concept that combines different methodologies. The ACES concept combines multiple models and methodologies into a single system where each component model or methodology fulfills a particular requirement of anthropometric casualty assessment. We defined the initial ACES concept to consist of the following six component models:

  1. 1.

    DHM – 3D model of the human form.

  2. 2.

    DEM – 3D model of equipment/PPE.

  3. 3.

    Penetrating Projectile Model – Represents impacts by penetrating projectiles (e.g., fragments, bullets, pellets).

  4. 4.

    Ballistics Material Model – Calculates the critical velocity required to penetrate the material and the residual velocity of the projectile.

  5. 5.

    Translation to Casualty Assessment – Translates from the DHM to casualty body model.

  6. 6.

    Casualty Assessment Methodologies – Quantifies harm caused by penetrating projectiles.

By connecting these low-level models and software components into a single package, the ACES concept will be a comprehensive tool for analysts. A depiction of the components of the ACES concept can be seen in Fig. 2.

Fig. 2.
figure 2

Components of the ACES concept

Full size image

4 Analysis Methodologies to Access Fit, Form, Wound Severity, and Level of Protection for Anthropometric DHM and DEM

In order to support the analysis process, we researched and developed methodologies and algorithms to perform the following tasks:

  1. 1.

    Clean the mesh, fill gaps, and reconstruct missing areas of scan data to ensure watertightness.

  2. 2.

    Rig and weightFootnote 1 the DEM to model the characteristics of sizing components used to fit the DEM to the DHM.

  3. 3.

    Rig and weight the DHM to support posing the DHM into operational postures (e.g., prone, kneeling, standing).

  4. 4.

    Detect intersection of projectile with DEM, determine material type at point of impact, and use material type to calculate critical and residual velocity.

  5. 5.

    Determine intersection and entry angle of projectile with DHM and associate with Sperrazza-Kokinakis (SK) casualty estimation body region (Kokinakis and Sperrazza 1965).

We researched and developed methodologies for converting full-body laser scan data to create DHMs. We investigated the conversion workflow by converting several scans from the Civilian American and European Surface Anthropometry Resource (CAESAR®) data set. (CAESAR® data currently is being used because it is commercially available). The methodologies created in this R&D will be applicable to ANSUR II data. We focused the research on investigating existing methodologies for converting digital human laser scans to DHM suitable for anthropometric casualty assessment, identifying technology and research gaps, and analyzing competing methods and technologies. An example of a finished scan is presented in Fig. 3.

Fig. 3.
figure 3

Cleaned, Decimated Scan: shows the scan in its final state prior to rigging. The scan is free of all geometrical errors, decimated, and watertight.

Full size image

We used digital artists to investigate a mix of available automated and manual tools used for converting scans to DHM and documented the process steps requiring special tools and/or manual conversion. We identified the following process steps as requiring digital artist manual intervention: (1) find and remove non-manifold edges; (2) find and remove artifacts; and (3) find and remove intersections and overlapping faces. We identified the methodology for preparing DHM scan data as consisting of three critical process steps: mesh decimation, mesh simplification, and mesh sealing. Again, we used digital artist to perform these process steps using three sets of anthropometric data selected from data that had been collected from previous projects.

The initial merged scan data produced an extremely dense, non-uniform mesh containing approximately 300,000 to 600,000 polygonal faces. This data was unnecessarily large for both visualization and simulation purposes, so the digital artists loaded the mesh into Meshlab (MeshLab 2012) to begin the optimization process. We selected MeshLab because it offered several of the required mesh processing algorithms. TSE digital artists first decimated the scan using the quadric edge collapse decimation tool to a target size of 20,000 polygonal faces, creating a less dense and therefore more workable mesh. The quadric edge collapse decimation tool works by specifying the target amount of polygonal faces for the final mesh, allowing options to preserve surface area and form after decimation. This allowed the final mesh to have fewer polygonal faces while still accurately representing the original form. Figure 4 includes an image of the original scan and the processed DHM, which is indicative of how the original form was preserved.

Fig. 4.
figure 4

A comparison between the feet of the original scanned file and the scan in its final state. On the left, the scan is of an overly dense resolution and full of holes and intersections. On the right, the intersections are cleaned up, and all of the holes in the mesh are filled.

Full size image

The scans were re-topologized using the surface reconstruction algorithms with Meshlab and resurfacing to create a uniform mesh. By the end of this step, the mesh holes and irregularities were corrected so the resulting form was as close as possible to the original scanned form. A before-and-after example of this process is shown in Fig. 4. We also researched methodologies for rigging and weighting DHM and evaluated the capabilities of Commercial Off-The-Shelf (COTS) software solutions such as the 3ds Max software application, which contains prebuilt rigs that could be used on scan data.

The rigged and weighted DHMs could be posed into different postures. We posed the three DHMs in operational postures and positioned the scans into a set of operational poses, including armed standing, kneeling, and prone postures (illustrated in Fig. 5). Digital artists made continuous alterations to the positions of the clavicle, pelvic, spine, and limb controls until realistic poses were achieved and used orthographic and perspective views to increase the efficiency of the process.

Fig. 5.
figure 5

A rigged DHM in an armed prone posture (left), armed kneeling posture (middle), and armed standing posture (right).

Full size image

The DEM is a 3D representation of the equipment form used to determine if a piece of equipment can fit a given DHM. It determines where and how well the DHM is protected and if a projectile traveling along a certain vector will strike the equipment. The DEM must meet the following requirements:

  • Simulate adjustments realistically to achieve a better fit, such as pulling or loosening straps.

  • Support testing for geometric intersection by a ray and report the location of all intersections.

  • Be watertight.

TSE digital artists created a digital model of the IOTV and ACH as a use case to understand the process of PPE rigging and enabling flexible elements for fitting to the DHM. These were modeled using photographic references and the artist’s firsthand knowledge and were rigged to enable flexible elements to bend realistically in order to achieve realistic fit. Subject matter experts examined the final DEM to ensure the characteristics of the equipment and how it was fitted to an individual was maintained.

We researched physics-based methods for simulating clothing dynamics for modeling equipment fit and investigated existing technologies, including the cloth solver in Blender (Blender 2012), a 3D modeling program, and the soft body physics capability provided by the Bullet Physics engine. Our analysis of these technologies indicates that while they provide realistic draping behavior over a human model, they assume a uniform, non-elastic material type. Because PPE typically is a combination of flexible, rigid, and adjustable components, the usual assumptions of clothing models are inadequate. We concluded that a more sophisticated simulation approach should be identified or created for use as a tool to investigate clothing dynamics and model fit. We investigated and evaluated options for modeling clothing dynamics and determined the level of work required to implement a clothing dynamics model was extensive. Alternatively, we developed a methodology using standard rigging and weighting of DEM in such a way that it could be fitted realistically. Three DHM poses, fitted with the IOTV, are shown in Fig. 6.

Fig. 6.
figure 6

A DHM equipped with PPE in each posture – the end result of the fitting process

Full size image

5 Conclusions and Future Work

This paper describes the ongoing research and development of approaches for analyzing fit and form of PPE, specifically body armor systems, to better account for a range of individual body shape differences and enhance protection. Future work on the project will include the investigation of commercial applications and extending capabilities to reach a broader community of users such as law enforcement, firefighting, and other hazardous occupations.