ABSTRACT
This effort, motivated and guided by prior simulated injury results of the unprotected head, is to assess and compare helmet pad configurations on the head for the effective mitigation of blast pressure transmission in the brain in multiple blast exposure environments.
A finite element model of blast loading on the head with six different helmet pad configurations was used to generate brain model biomechanical responses. The blast pressure attenuation performance of each pad configuration was evaluated by using the calculated pressure exposure fraction in the brain model. Monte Carlo simulations generated repetitive blast cumulative exposures.
Significant improvement of a 6-Pad Modified configuration compared to a 6-Pad Baseline configuration indicates the importance of providing protection against the side blast. Both 12-Pad configurations are very effective in mitigating pressure in the brain. Repetitive blast exposure statistics for operational exposures shows that pad configurations with a larger number of pads and smaller gaps between pads perform better than the configurations with a smaller number of pads and larger gaps between pads.
Optimizing helmet pad size and/or placement could provide an improved protection by minimizing the side blast orientation effects and mitigating high-pressure fields in the brain from repeated blast exposures.
INTRODUCTION
Modern combat helmets are a combination of stiff ballistic polymer composites and compliant energy-absorbing foam materials designed to protect the head from traumatic brain injury (TBI).1,2 The head itself is a combination of stiff and compliant materials, including the external soft tissues, hard cranium, and soft brain tissues. Blast pressure transmission through the head-helmet system into the brain is a complex dynamic event modulated by the combat helmet system.3,4 Blast sensors promise to characterize these blast pressure environments and provide better injury diagnosis and risk manage.5
Multiple blast exposures, in operational or training environments, can be expected to feature different blast pressure magnitudes and different incident directions of the incoming wave.6 Each blast generates different internal distributions of dynamic pressures in the brain. This complicates understanding of the cumulative effects of multiple exposures, the risk of brain injury, and establishment of injury criteria. The utility of blast sensor information will be maximized by a better understanding of how the transition from a “low” to “high” number of blasts occurs from the statistics of these dynamic internal brain pressures. Various computational models have been used to establish the relation between the blast pressure on the head and helmet and the biomechanical response in the head, increase understanding of blast injury mechanisms, assess helmet performance, and identify promising helmet designs.7–9
In our earlier work,10 we generated a brain dynamic pressure response database of different incident blast overpressure and orientation combinations for the unprotected head and used it for Monte Carlo analyses of random sequences of blast events. The statistical range (of the internal brain pressure response metrics versus number of blast) transitions from varied responses at “low” numbers of blasts to a more convergent response at “high” number of blasts. The cumulative pressure exposure fraction (CPEF) quantitatively characterizes the differences in notional operational and training exposures.
The objective of this effort, motivated and guided by the simulated injury results of the unprotected head, is to assess and compare different helmet pad configurations on the head to effectively mitigate blast pressure transmission in the brain for multiple blast exposures. Several hundred biomechanical simulations of dynamic pressure response in the brain are performed to generate a database of representative results for blast pressures and incident directions in all pad configurations. The database is queried for the Monte Carlo analysis of longer blast sequences representing operational exposures to random blast overpressures and orientations. The statistics of the internal brain pressure responses are then analyzed to identify and quantify more optimal pad configurations.
METHODS
Blast pressure transmission into the brain occurs by a combination of direct exposure of the head, infiltration under the combat helmet between the pads, and propagation through the combat helmet. Direct simulation of individual blast event constituting training or operational blast sequences, using three-dimensional blast, head and helmet fluid-structure interaction models, is computationally expensive.11 Simulation of random sequences of blast events would be prohibitive. The approach taken here uses a combination of two-dimensional detailed computational simulations, probabilistic calculations using a database constructed from the detailed results, and statistical analysis of the data generated for blast sequences. This is more efficient in terms of computational resources and time required to generate the results, demonstrate the methodology, and provide key insights.
Computational Finite Element Model of Human Head with Pads
As previously described,10 we used a transverse 2D section (axial plane) from a 3D high-resolution human head model12,13 subjected to the blast loading. The quadrilateral mesh, with an average element size of 0.5 mm, consists of 5 different structures, that is, gray matter, white matter, ventricles, cerebrospinal fluid, and skull, as shown in Fig. 1A. The gray matter and white matter were modeled as isotropic viscoelastic materials with slightly different constants. The cerebrospinal fluid in the layer between the skull and brain and in the ventricles inside the brain was modeled as a hyper-elastic solid with a very low shear modulus to capture the fluid-like behavior. The skull was assumed linear elastic as a result of the small strain experienced during blast.
(A) A 2D slice in transverse/axial plane of head. (B) Range of incident overpressure and orientation to head in operational blast exposure. (C) Configurations of No-Pad, 6-Pad Baseline, 6-Pad Modified, 5-Pad, 2-Pad, 12-Pad A, and 12-Pad B used for the comparison of protective performance. (For the color figure, the reader is referred to the web version of this article.)
The ConWep model14 is used to apply the transient blast loading on the head resulting from a spherical free-field explosion in air (without the ground effect). The bare high explosive charge is detonated near the head surface and with an angle of θ relative to the middle sagittal plane of the head. Fig. 1B shows the range of incident blast overpressures and orientations used for a notional repetitive blast exposure environment. Four explosions with increasing trinitrotoluene weights at a stand-off distance of 2.7 m are considered. The resulting peak incident overpressure ranges from 7.5 psi (51.7 kPa) to 15 psi (103.4 kPa) with the increment of 2.5 psi (17.2 kPa). The angle of θ ranges from −180 to +180 degrees with the increment of 15 degrees. The notional “random” operational/combat exposure environment covers all 4 incident overpressures and all 24 angles from −180 to +180 degrees.
Based on the simulated injury results from the unprotected head, different configurations of helmet pad suspensions are explored to compare the blast load transmission into the brain. As shown in Fig. 1C, we used six different pad configurations labeled 6-Pad Baseline, 6-Pad Modified, 5-Pad, 2-Pad, 12-Pad A, and 12-Pad B to compare the protective performance in the random operational blast environment. To make a consistent comparison, we fix the total pad contact length in these 2D configurations as 14 inches and place the pads on the head with left-right symmetry. It is assumed that the pads absorb all the incoming blast loading, and thus, the blast pressure is only applied to the area between the pads on the head in each pad configuration.
From the prior results on the unprotected head model, the blast orientation of ±75° predicts more severe brain injury than other blast orientations.10 This influences the approach taken to assessing the effect of helmet pad. The 6-Pad Baseline configuration is based on a standard pad placement. For both 6-Pad Modified and 5-Pad configurations, 2 pads are centered along the ±75° lines, respectively. To assess how the size of individual pads affects the protection, we have also designed three other configurations of 2-Pad, 12-Pad A, and 12-Pad B having the same total contact length. The 12-Pad A configuration differs from 12-Pad B by rotating pads 15° along the head circumference.
The pressure-based TBI thresholds of 142 and 173 kPa suggested in the literature15,16 were used to assess the repetitive blast mild TBI (mTBI) and single blast mTBI, respectively. The result of each finite element (FE) simulation with time duration of 2.5 ms produced spatially and temporally resolved pressure field data for the whole head that can be compared to these pressure thresholds for respective injury criterion to determine if blast mTBI has occurred. The criteria are simultaneously applied on an element-by-element basis to the brain at every time step in each analysis. If an element’s maximum pressure exceeds a given pressure threshold, the element is considered to have been “injured.”
Monte Carlo Simulation of Repetitive Blast Sequences
We refer the reader to Reference [10] for further details. For this study, the time interval between two consecutive blasts is assumed to be longer than the so-called the window of increased cerebral vulnerability (WICV).17 It has been shown that if a second blast to the head takes place within the WICV, reinjury triggers a more severe, cumulative damage type of cell response relative to the initial biomechanical response.18 In this article, we do not consider such a short time interval effect on the viscoelastic brain properties and associated injury criteria used in the simulation of repetitive blast sequences.
Maximum dynamic pressures in brain model from single blast event of incident pressure P = 15 psi (103.42 kPa), angle of incidence = 0, 75, 135, and 180 degrees for all 7 configurations. (For the color figure, the reader is referred to the web version of this article.)
RESULTS
Pressure Response of the Head With Pads Subjected to Single Blast
Blast pressure transmission through the brain from the skull is a complex dynamic event. Fig. 2 shows the example simulation results of the maximum pressure field in the brain model for the blast with the highest incident pressure of 15 psi (103.4 kPa) and at angles of 0, 75, 135, and 180 degrees (a subset of the total 24 angles considered) for configurations of No-Pad, 6-Pad Baseline, 6-Pad Modified, 5-Pad, 2-Pad, 12-Pad A, and 12-Pad B. The results for lower incident pressures of 7.5 psi (51.7 kPa), 10.0 psi (68.9 kPa), and 12.5 psi (86.2 kPa) are qualitatively similar and not shown here. It was seen that the blast orientation significantly affects the pressure field, particularly for the configurations of No-Pad and 6-Pad Baseline, both showing the prediction of more severe brain injury around ±75° blast orientations.
As explained, the differences in pressure infiltration in the No-Pad configuration highlight the influence of local head curvature.10 Compared to the curvature of the anterior or posterior head, the sides of head are comparatively flat and more conducive to uniform pressure transmission into the brain. The pad placement on the head not only reduces the pressure in the brain but also changes substantially the pressure pattern in the brain when compared to the No-Pad configuration at the same blast orientation. In the six pad configurations, the large pressures appear in the coup region between pads where there is nothing to block the blast, while lower pressure occurs at the coup region where there is pad coverage. The pressure in the contrecoup region is much smaller in both amplitude and area except in the cases of No-Pad at 75 degrees and 6-Pad Baseline at 75 degrees.
These observations on the maximum pressure patterns for these orientations are significant toward understanding and quantifying repetitive exposure statistics for pressures exceeding injury thresholds. In Fig. 3, the CPEF parameter is plotted parametrically for a single 15 psi (103.42 kPa) overpressure (N = 1), all 24 orientations, and injury pressure threshold of 173 kPa for all 7 pad configurations. They are broadly symmetric and feature significant regions of both higher and lower CPEF values. For configurations of No-Pad and 6-Pad Baseline, orientation intervals between (−120, −60) and (+60, +105) degrees have the highest CPEF values, while (−45, +45) and (+105, −135) degrees have the lowest values. The slight asymmetries in the results are as a result of natural anatomical asymmetries present in the model. The intervals between (−135, −120), (−60, −45), (+45, +60), and (+105, +135) degrees are the strong transition regions between the lowest and highest CPEF values. The 6-Pad Modified configuration, which differs from the 6-Pad Baseline configuration by a 15-degree shift in pad placement, shows much smaller CPEF for all angles, especially in the significant reduction of high injury fraction at the side blast orientations seen in both No-Pad and 6-Pad Baseline configurations. The 5-Pad configuration has a high CPEF around 0 degree, while the 2-Pad configuration has a high CPEF around 180 degrees, highlighting the effect of anterior and posterior exposures. Both 12-Pad A and 12-Pad B configurations result in very small CPEF when compared to others, highlighting the positive influence of smaller pad gaps. Plots using the injury pressure threshold of 142 kPa, not shown here, display similar patterns but with higher CPEF values.
Cumulative Pressure Responses of the Head With Pads Subjected to Multiple Blasts
Using the database compiled from the 96 biomechanics simulations, from 4 incident blast overpressures and 24 blast orientation angles, Monte Carlo simulations were conducted for 20 random sequences of Ntot = 200 blasts for each of the 7 pad configurations to generate cumulative response exposures. The left column of Fig. 4 lists all seven pad configurations. The center column shows the CPEF contour plots exceeding the 142 kPa repetitive TBI threshold. The effect of pad configuration is significant, highlighted by the selection of CPEF axes to visualize the wide range of responses. It can be seen that for both No-Pad and 6-Pad Baseline configurations the predicted repetitive blast injury above a CPEF(e, Ntot) of 0.05 covers the entire brain. For the 6-Pad Modified configuration, the predicted injury occurs mainly in small areas between pads. The 5-Pad configuration has injury at the frontal lobes of the brain, while the 2-Pad configuration has injury largely at the occipital lobes of brain. There is essentially no injury seen in both 12-Pad A and 12-Pad B configurations.
Cumulative pressure exposure fraction (CPEF) versus blast orientation angles for single blast overpressure of 15.0 psi (103.4 kPa) (N = 1) and TBI threshold of 173 kPa for configurations of No-Pad, 6-Pad Baseline, 6-Pad Modified, 5-Pad, 2-Pad, 12-Pad A, and 12-Pad B. (For the color figure, the reader is referred to the web version of this article.)
The right column of Fig. 4 shows the envelope curve for each set of 20 CPEF(Etot, N) sequences exceeding TBI thresholds of 142 kPa in the brain along with curves of the cumulative mean and cumulative standard deviation (SD), respectively. The CPEF parameter is sensitive to the magnitude and direction of blasts early in Ntot = 200 sequence of exposures, converging as the number of blasts increases. Note that at 200 blasts, the envelope curve has not reached a single mean value but is contained within a gradually narrowing band of values around the mean. The number of blasts to reach a “practical” equilibrium state for the mean values and residual SDs for all seven configurations are listed in the figure. This equilibrium state is one in which each additional blast, while continuing to increase the cumulative exposure, does not significantly change these two quantities. Among all pad configurations, the CPEF mean value converges at the lowest number of blasts for the 5-Pad configuration. It converges at the highest number of blasts for the 12-Pad A configuration. The pad configurations, rated in the descending order of cumulative mean value after Ntot = 200 blasts, are: No-Pad (|$\sim$|0.17), 6-Pad Baseline (|$\sim$|0.13), 2-Pad (|$\sim$|0.008), 5-Pad (|$\sim$|0.0035), 6-Pad Modified (|$\sim$|0.0022), 12-Pad B (|$\sim$|1.5 × 10−6), and 12-Pad A (|$\sim$|1 × 10−7).
DISCUSSION
The results presented earlier provide insight into the way individual blast events expose the brain to higher pressures exceeding TBI thresholds for seven different pad configurations. This information is then used to describe how multiple blast event exposures are accumulated in the brain. Different aspects of this study reinforce each other and quantify these trends.
First, the maximum pressure encountered at each point in the brain model from an individual blast creates complex spatial patterns across the brain. The patterns vary strongly with both blast orientation and pad configuration for the same blast overpressure. The brain injury areas with pads on the head are reduced when compared to the No-Pad configuration. In both No-Pad and 6-Pad Baseline configurations, blast from around the ±75° orientations results in larger brain model injury areas than blast from other orientations because of the effect of the low local skull curvature. Conversely, the pad configurations covering the ±75° location on the head result in significantly smaller injury areas in the brain. The improved protective performance by smaller pads indicates that minimizing large gaps between pads is of key importance. The coup and contrecoup injuries seen in the No-Pad configuration become less severe with pads on the head. The injury area is more concentrated at the coup region when there is no pad coverage in the direction of blast pressure. The injury area at the coup becomes smaller when the corresponding gap between pads shrinks.
Second, the CPEF quantifies those portions of the brain exceeding TBI thresholds from multiple blast exposures for all pad configurations. It can track multiple blast exposures at each location in the brain and exposure patterns across the brain. It captures the complex geometric pattern overlays that develop from different blast orientations, highlighting local regions of high and low cumulative exposure. It confirms that, in the blast environment, the low local skull curvatures on the left and right sides make the brain more susceptible to injury in both No-Pad and 6-Pad Baseline configurations. Furthermore, carefully designed pad configurations can markedly relieve the adverse skull curvature effects for multiple exposures. It is also possible to counteract the blast orientation effect and improve the overall blast protection capability by judicious pad placement. A larger number of smaller pads, having smaller gaps between pads, can potentially achieve better brain protection than a smaller number of larger pads with the same pad coverage.
Example of Monte Carlo results for operational random sequence. Center column: contour of cumulative pressure exposure fraction (CPEF) (e, Ntot), Ntot = 200 in brain model, for seven pad configurations. Right column: envelope curve of 20 CPEF (Etot, N) sequences exceeding repetitive TBI thresholds of 142 kPa in brain, along with curves of cumulative mean value and cumulative standard deviation (SD). The figure lists the CPEF converged values and associated number of blasts n for the cumulative mean and cumulative SD. (For the color figure, the reader is referred to the web version of this article.)
Third, the Monte Carlo simulations generate the evolution of the CPEF, the cumulative mean, and cumulative SD versus number of blasts N for diffident pad configurations. The evolution of CPEF quantifies the relative pad performance and identifies the optimal pad configuration in a more precise fashion. The CPEF itself is most sensitive, as shown by its fluctuations and range of responses, at lower N. The cumulative mean value is a more global measure of pressure “dosing.” Its convergence between approximately 40 and 70 blasts means that in the range of 0 to 40 blasts the evolving CPEF pattern in the brain model is changing in both the geometry and magnitude. The cumulative SDs converge between the first 80 and 120 blasts meaning that in the range of 40 to 80 blasts the geometric details of the evolving patterns in the brain model will still be important and changing although the average total exposures will be similar.
This article highlights a problem and presents an approach to developing quantitative recommendations for military medicine by simulation and physical measures. These include (1) characterizing the protective role of helmet pad suspensions in the anticipated operational scenarios, (2) fusing intelligence and operational planning for area-specific threats, and (3) software prediction of CPEF ranges and convergence thresholds to guide forward command and medical decisions.
The assumptions and limitations of this work are summarized here in the remaining portion of the discussion. We made the assumption of pads absorbing all the incoming blast loading on the exterior of the helmet based on prior observations and considerations. The blast pressure does load the exterior of the helmet, transmitting stress through the pads and onto the head surface. The contact pressures between the energy absorbing pads and the head, however, are considerably lower and of longer duration compared to the higher blast pressures of shorter duration directly on the head under the helmet (or directly on the exposed face and neck area of the head).19 Because of these comparatively low stresses of longer duration, the force on the head through the pads was not included. The blast pressure case is very different, of course, compared to the blunt impact case, where all forces affecting head response are indeed transmitted onto the helmet and through the pads.20 Our prior and ongoing efforts quantify how blast affects a 3D head model with a helmet and how stresses are attenuated by the strain-rate dependent pad suspension properties.21
The simulations were based on a 2D model of a transverse plane of the head. The helmet at the same transverse plane was not considered because the head model would have been fully protected by the helmet and not experience any blast pressure exposure. Using a 3D model of the head and helmet with various pad configurations, we will be able to simulate the blast pressure directly on the helmet, infiltration under the helmet and between the pads, and directly on the exposed face and neck.22
The injury criteria and threshold value for the repetitive blast mTBI was established based on the data in the literature.15,16 We used the same criteria for both gray matter and white matter in the brain. Tissue- or region-specific injury criteria, including changes as a function of prior exposures, can be incorporated into this type of analysis when a consensus develops in the literature.
CONCLUSION
A practical, systematic, and quantitative computational approach to evaluating and optimizing helmet pad placement to mitigate repetitive blast pressure infiltration into the brain is developed and demonstrated. It provides information that complements and can maximize the utility of data from blast sensors. It generates results for random blast sequences using a reference database, generated by a set of detailed biomechanical simulations of blast on the head with pads, and employing the Monte Carlo method. Statistical convergence of the internal brain response metrics versus number and type of blast exposures quantifies the blast mitigation by judicious placement of pad suspension components on the head to alleviate blast pressure infiltration in areas of low skull curvature. This process is computationally efficient and extendable from this 2D demonstration to a 3D implementation.
ACKNOWLEDGMENTS
The authors acknowledge Dr. Amit Bagchi from NRL for technical discussions.
FUNDING
This study was supported by the Office of Naval Research (ONR) through the Naval Research Laboratory’s Basic Research Program (Work unit: 63-6A72) and the Department of Defense (DOD) High Performance Computing Modernization Program (Project: NRLDC04161572) at the Army Research Laboratory, Air Force Research Laboratory, and Army Engineer Research Laboratory DOD Supercomputing Resource Centers.
REFERENCES
Author notes
Presented as a poster at the 2019 Military Health System Research Symposium, Kissimmee, FL; MHSRS-19-1406.
The views expressed in this article are those of the authors and do not necessarily represent the official position or policy of the U.S. Government, the Department of Defense, or the Department of the Navy.
This material is declared a work of the U.S. Government and is not subject to copyright protection in the United States. Approved for public release; distribution is unlimited.
