Protective Headgear Attenuates Forces on the Inner Table and Pressure in the Brain Parenchyma During Blast and Impact: An Experimental Study Using a Simulant-Based Surrogate Model of the Human Head

Rates of traumatic brain injury (TBI) among warfighters have been rising during recent conflicts and this injury has a profound impact on the health of combat veterans [1–6]. For soldiers deployed to Afghanistan (operation enduring freedom) and Iraq (operation Iraqi freedom (OIF)), it has been estimated that nearly 25% of all deployed warfighters suffered from symptoms of TBI [7–9]. Blast exposure accounts for the majority of total injuries among contemporary military personnel, with reports as high as 80% [10–13]. It is estimated that 50–60% of head injuries involving blast exposure on the battlefield also involved blunt trauma [14]. During the operation enduring freedom and OIF conflicts, the use of explosive devices increased, with improvised explosive devices becoming the weapon of choice against coalition forces [15], resulting in increased blast exposure for the modern warfighter [16,17]. The U.S. Navy-Marine Corps combat trauma registry reported that improvised explosive devices were found to be the most common mechanism responsible for TBI cases (52%) during OIF [18]. Effective mitigation strategies are strongly desired by Canadian and U.S. militaries, but in order to reduce the occurrence of blast-induced traumatic brain injury (bTBI), a proper understanding of the bTBI injury mechanisms is required [10].

A blast event can cause many different types of injury, which are broadly classified into four categories [10,19]. Primary blast injury is due to the direct effects from the blast wave propagating through the body and damaging living tissue. Secondary blast injury involves the propulsion of debris or fragments that can cause blunt or penetrating injury. Tertiary blast injury occurs when the explosion induces global acceleration of the body, propelling it into collision with surrounding structures in the environment. Quaternary blast injuries are due to postdetonation environmental stressors, such as chemical exposure or burns from thermal radiation. There is limited information available on the effect of primary blast on the human head [10,13], and it is not yet clear how significant primary blasts are relative to blunt impact as sources of injury [14]. Due to the sensitivity of brain tissue to loading rate [20], the mechanisms of injury for primary bTBI may not be similar to conventional blunt impact induced TBI [13,21,22]. In this work, we created primary blast exposure using free-field detonation experiments and also simulate impact to the head, which could arguably simulate a tertiary blast injury scenario involving direct head impact [23].

The mechanisms of primary bTBI remain poorly understood, however several theories exist [10,24]. One theory suggests that blast waves can damage brain tissue by traveling through skull orifices such as nasal sinuses [25], auditory canals [26,27], and/or eye orbits [28–31]. It has also been suggested that a blast wave could directly transmit into the brain through the skull via transosteal wave propagation, referred to as the direct cranial transmission theory [32–34], however acoustic impedances suggest that this is an unlikely mechanism [10]. Another theory suggests that skull deformation can be induced from a blast wave [35–37], either through localized deformation or by exciting a mode of skull resonance. These skull deformations can then cause significant intracranial pressure (ICP) fluctuations [10,38]. Another theory referred to as thoracic surge is proposed to cause bTBI indirectly through the rapid compression of the thorax and abdomen from the blast wave, which then transmits a pressure wave toward the brain through soft tissue and vascular mediums [39–41]. Cavitation has also been proposed as a primary blast injury mechanism, in which localized regions of negative ICP may result in the formation of cavitation bubbles that cause damage to surrounding brain tissue when they collapse [42,43]. These theories on primary bTBI remain under dispute, but likely do not occur in isolation from one another during a realistic blast scenario, and may vary based on the loading magnitude and exposure duration [10,24]. Understanding which mechanisms are most prevalent during a blast scenario and how they can effectively be mitigated could help direct the development of future protective combat headgear.

The mechanisms of impact induced head injury [44] are relatively better documented than are the mechanisms of blast injury. Impact forces on the head create abrupt head motions comprising both linear and rotational kinematics. These impact forces are first applied to the scalp and then skull, which may, in turn, create relative motion between the skull and the brain within the cranial vault. These relative motions are believed to be a possible cause of disruption to bridging veins (usually implicated in subdural hematoma) [45,46] and also of contact between the brain and skull. Impact to the head and brain also elicit accelerations that are both linear and rotational which are thought to result in inertial loading of tissue leading to both local (focal) and widespread (diffuse) tissue damage [47]. Impact induced injuries are generally classified as either focal or diffuse. Examples of focal injuries include contusion, laceration and localized skull fracture, the latter of which is typically considered to be caused by localized forces on the skull and has in a large amount of literature been correlated with linear head acceleration following impact [48–50]. A commonly referenced diffuse injury involves widespread axon disruption, the severity and risk of which has been correlated with rotational impact mechanics including acceleration and velocity and is sometimes discussed in the context of brain injuries including concussion [47,51–53]. Often, after a head impact, both linear and rotational kinematics result and therefore it is possible to have both head and brain injuries that are focal and diffuse, affecting both the skull and brain.

Helmets are the primary protection device fielded to protect the head of the wearer from impact, as well as ballistic threats for military personnel, and they are commonly acknowledged as being effective in protecting the wearer from these threats [14,54–60]. In combat scenarios, no existing protective headgear is able to fully protect against all the potential threats that can exist on the battlefield [56]. Current combat helmets are tested by their ability to prevent ballistic penetration, mitigate blunt trauma, and behind-helmet blunt trauma from projectiles and by their ability to withstand environmental stressors such as exposure to temperature extremes [14]. Ballistic protection of the head is achieved by the helmet exterior shell that is struck by projectiles, and through a process involving deformation and failure of shell materials, the exterior shell redistributes the loading over a larger area and reduces the amount of energy available for transmission to the helmeted head. These reductions in energy transfer are typically inferred from head models that use clay to measure depth of deformation of the helmet interior [61], and more recent proposals use head models capable of measuring forces exerted on the head exterior during helmet deformation. Reduction in energy transmission to the underlying head is generally associated with reduced risk and severity of head injury. Similarly, for impact protection, helmets are tested for their ability to attenuate impact transmission to an underlying head [14], typically quantified by measuring linear head acceleration during the impact event. A helmet that leads to lesser head acceleration than that measured with another helmet is typically thought to offer superior impact attenuation, and head protection [62]. It is evident from the above that the paradigm for assessing head protection is quantifying reductions in mechanical measures. Despite the efficacy of contemporary combat helmets in mitigating injury from ballistic threats and impacts, bTBI rates continue to rise [63], and it appears to be an open question whether modern combat helmets are effective at protecting military personnel from blast waves. Indeed, there is scant experimental work that assesses whether or not contemporary military head protection is capable of altering the transmission of blast energy into the head.

The objective of this study is to ascertain whether or not protective headgear alters the transmission of blast and impact energy into the brain, using a biofidelic surrogate of the head based on simulant materials. Free field blast simulations at an outdoor test range were conducted to create realistic blast loading on the head surrogate for cases where the head was unprotected, wearing a contemporary military helmet, and wearing both a helmet and eye protection. Blunt impact loading on the head surrogate was also created in a laboratory setting for the cases where the head surrogate was unprotected and protected using a helmet. Within the surrogate head, forces on the inner table of the skull were measured, along with pressures within the brain parenchyma. Experimental outputs (measurements) were peak measures of force and pressure as well as quantification of frequency components in the force and pressure data. Descriptive statistics of repeated experiments were quantified for the purpose of comparing measurements between cases of unprotected and protected headform scenarios.

For this work, we instrument a surrogate blast headform known as the blast injury protection evaluation device (BIPED), fully developed by Defense Research and Development Canada (DRDC). An in-depth report on the BIPED, including the history of development, along with current headform geometry and physical properties, is presented in the previous literature [64]. The BIPED has been subjected to simulated blast overpressure in blast tube experiments and has been shown to measure ICP of comparable peak magnitude, time duration, and frequency content to cadaver heads that were subjected to overpressure in the same blast tube [65]. In this paper, we outline key headform properties and provide details on the instrumentation used for this study.

The BIPED, shown Fig. 1, has several internal and external biofidelic features such as anthropomorphic eye orbits, nasal cavities, a brain, cerebrospinal fluid, skin, a jaw, ears, a cerebellum cavity, along with falx and tentorium membranes. Surrogate skull material is nominally 6.35 mm thick and composed of TC-854 A/B polyurethane (BJB Enterprises Inc., Tustin, CA). Surrogate brain tissue (exposed in Fig. 1(a)) is composed of the silicone Sylgard 527 (Dow Corning Corporation, Auburn, MI), and water is used to simulate cerebrospinal fluid. The external soft tissue layer (Fig. 1(a)) is composed of a tough urethane rubber (Vytaflex 20, Smooth-On Inc., Oroville, CA). Internal falx and tentorium membranes are modeled using 70 A hardness Neoprene layers. External feature locations and size are based upon head models used in, for example, Canadian Standards Association CAN/CSA Z262.6-02, helmet test standards [66]. The upper skull geometry matches the top half of the ISO Size J Headform (Cadex Inc., QC, Canada), and internal skull features are based upon an open-source CAD model of the human skull [67]. The BIPED has a four-point attachment location at the base of skull where it can be mounted onto a Hybrid III neck.

For the experiments performed in this study, a BIPED is instrumented with several pressure transducers, similar to work done previously [68,69], but it is additionally equipped with novel force transducers within the surrogate skull material. X-ray images showing some of these transducer locations can be seen in Fig. 1. In our previous work, in-fiber Bragg grating (FBG)-based optical force transducers were developed for measuring normal forces on the BIPED inner table, with frequency content characteristic of blast [70], and for this study, multiple FBG-based force transducers are calibrated for normal forces and integrated into the surrogate skull material and mounted flush with the inner surface (inner table) of the skull. The transducers are not calibrated to measure in-plane or shear forces. These transducers are composed of an FBG embedded within an aluminum and acrylic based multilayered cylindrical superstructure, with an overall size of 3 mm in height and 6.3 mm in diameter. The approximate transducer plane for which these transducers are integrated is shown in Fig. 2(a). This transducer plane is parallel to the transverse plane and offset by approximately 20 mm above the eye sockets on the surrogate skull. Individual transducer locations are shown in Fig. 2(b), with four locations for front, right, left (impact only), and rear transducers. Two surrogate skull caps are instrumented with an array of force transducers (one for blast experiments and one for impact experiments). After performing the blast experiments, the BIPED skull cap was replaced with another instrumented skull cap for impact experiments with an additional force transducer on the left side. The transducers are fixed into the headform using high-strength epoxy (Loctite MS 930, Henkel AG & Co., Dusseldorf, Germany).

In addition to the force transducers, the BIPED used for this study is instrumented with several pressure transducers for measuring internal pressures within the surrogate brain parenchyma, as well as the external pressure profile. Three Kulite piezo-resistive pressure transducers (XCL-072, Kulite Semiconductor Products, NJ, modified by the manufacturer to meet frequency response requirements) are used to monitor ICP. These transducers have a linear output for pressures up to 14 bars, a resonant frequency over 550 kHz, and are relatively insensitive to acceleration (less than 0.0001% full scale output per g force). These ICP transducers are molded into the surrogate brain material (locations shown in Fig. 2) in approximately the same plane as the force transducers. External pressure measurements are collected by pressure transducers (LL-125 flat line series, Kulite Semiconductor Products, NJ, modified by the manufacturer to meet frequency response requirements) located on the front, side, and back locations of the headform (shown in Fig. 2). These external pressure transducers were retrofit into the surrogate skin material and mounted to be flush with the outer surface of the BIPED. The BIPED is filled with water to model CSF and then sealed using thread sealing cord (Loctite 55-10, Henkel AG & Co., Germany) and self-fusing silicone (Loctite SI 5075, Henkel AG & Co., Germany) for blast experiments. For impact experiments, the BIPED was sealed using an industrial silane adhesive (Loctite MS 930, Henkel AG & Co., Germany).

The instrumented BIPED described in Sec. 2.1 is exposed to blast overpressure at a free-field testing facility located in Valcartier, QC, Canada. For these experiments, a 5 kg cylindrical C4 charge is detonated at a height of 1.5 m above the ground. The BIPED is positioned 5 m away horizontally, with the nose of the headform positioned at approximately the same height as the charge. Unprotected and protected side views of the BIPED can be seen in Fig. 3. For helmeted cases in this study, we use a tactical ballistic helmet (TBH-II HST Size Large, Gentex Corporation, PA). This helmet meets the modified and abbreviated NIJ Standard 0106.01 [71] for ballistic helmets using the NIJ Standard 0108.01 threat level IIIA [72] (9 mm 124 gr FMJ RN at 1400 fps). It also meets the modified and abbreviated U.S. Army ACH Helmet CO/PD-05-04: 2007 specification. The comfort liner of the TBH-II helmet was unaltered for experiments in this study. The same helmet configuration and location was used in all cases involving the TBH-II.

For blast experiments, the BIPED is fixed upon a hybrid III neck (Humanetics Innovative Solutions, MI), and the base of the neck is fixed upon a rigid vertical steel pole that is held secure to a concrete pad at the ground. The steel pole also helps to protect wires and optical fibers running external to the headform. ICP and external pressure signals were collected at 1 MHz and force signals on the inner table of the simulant skull were gathered at 40 kHz, using a commercial interrogator (SmartScan, Smart Fibers Ltd., UK). A schematic of the test platform can be seen in Fig. 4, in which the BIPED is facing the charge (face-on).

Four blast experiments are repeated using an unprotected (bare) BIPED, and five experiments are conducted using various headform protection scenarios, for a total of nine blast experiments and three protection scenarios. The different test scenarios for the nine blast experiments are outlined in Table 1. For the protected scenarios: three trials were performed using a TBH, and two trials were performed using the TBH and a visor (Baltskin Cobra, Revision Military Ltd., VT). After each blast experiment, the BIPED was visually inspected for obvious damage and also to verify that the simulated CSF volume was not changed due to fluid escape.

An instrumented BIPED (identical to the one used for blast experiments except lacking external pressure transducers, and that has an additional force transducer on the left side) is subjected to blunt impact experiments using a custom-built linear guided monorail drop tower [73,74]. All impacts are to the high forehead of the BIPED. High-speed cameras (Phantom v611, Vision Research, Wayne, NJ) are used to track global headform movement at 1000 frames per second. A soft impact surface consisting of sportsfield turf (SouthWest Greens GB-011, 100% polyethylene, 1.75 in pile height) with a crumb rubber infill was fixed upon a steel anvil at the base of the vertical drop tower. This soft impact surface was used instead of the bare steel anvil in order to protect the bare BIPED headform from potentially rupturing and leaking internal fluids at higher drop velocities. The BIPED is mounted onto a Hybrid III neck, which in turn is fixed to an adjustable gimbal that moves freely along the vertical axis of the drop tower. A still-image of a protected and unprotected BIPED drop tower setup can be seen in Fig. 5.

The height in which the headform is dropped from is controlled to achieve three tiers of impact velocity, termed “low” (1.47–1.54 m/s), “medium” (2.31–2.37 m/s), and “high” (3.25–3.33 m/s) to simulate a range of speeds that are plausible for a head falling from a height associated with someone that is kneeling. For each velocity tier, five drop experiments were repeated for an unprotected BIPED and a protected BIPED (using the same TBH model that was used for blast experiments), totaling 30 impact experiments. A test matrix for the BIPED impact experiments can be seen in Table 2. After each impact experiment, the BIPED was visually inspected for obvious damage and also to verify that the simulated CSF volume was not changed due to fluid escape. For these impact experiments, ICP data are collected at 25 kHz and force data is gathered at 20 kHz.

Data for this study is postprocessed using matlab R2016b (MathWorks Inc., MA). For the instrumentation used during blast experiments, including external pressure, ICP, and force signals, the raw data is filtered using a fourth-order Butterworth low pass filter with a cutoff frequency of 10 kHz, based on CFC 6000 [69,75]. Data from impact experiments is filtered using a fourth-order Butterworth low pass filter with a cutoff frequency of 1650 Hz (CFC 1000). This filtering approach uses filter specifications described by the Society of Automotive Engineers (SAE) Standard J211 for processing head kinematics and head/neck forces in impact experiments [76]. We acknowledge that the SAE standard does not discuss filtering approaches for forces on the inner table or parenchyma pressures. Nevertheless, our impact measurements indicate frequency content well below CFC1000 cutoff frequency; therefore, we feel this filtering approach is appropriate because it will not remove mechanical energy at key frequencies in our signals.

For blast experiments, pressure signals are all time synchronized and triggered by the signal sent to the detonator. Force signals are collected on a separate data acquisition system triggered manually just before detonation. During impact experiments, ICP and force signals are also collected on separate data acquisition systems. For comparison of data from different blast trials, signals are aligned based on the incident peak of the front external pressure transducer. For comparison within trials, force signals are aligned by shifting the initial peak of the front force transducer to be approximately halfway between the front external incident peak and the front ICP initial peak. For example, in many blast trials the front external pressure incident peak occurs at approximately 2.6 ms and the front incident ICP peak occurs at approximately 2.8 ms; therefore, the force signals are shifted such that the front initial force peak occurs at 2.7 ms in this instance. We acknowledge that this may not be the exact time where the force signal peaks would occur, however, for the visualization purposes in this work, the error from using this synchronization method should be no greater than 100 μs and the true difference would not be distinguishable for the figures shown in this paper. In addition, our subsequent analysis focuses on magnitudes of measured kinetics as opposed to their relative order of sequence in the time domain, and so we believe our data analysis approaches are justified. Consequently, any conclusions drawn are limited to magnitudes of measured kinetics.

For spectral analysis, we use a fast Fourier transform (FFT, implemented using matlab R2016b) to investigate whether the various protection scenarios have an effect on frequency content [24] in measured data. We also examined the accumulation of amplitudes with increasing frequency using normalized cumulative FFT data, which contain the cumulative sum of FFT amplitudes normalized by the total value of all amplitudes accumulated up to the Nyquist frequency for the given blast or impact experiment. For blast experiments, the Nyquist frequencies are 250 kHz (pressure data) and 20 kHz (force data). For impact experiments, they are 12.5 kHz (pressure) and 10 kHz (force). These normalized data convey the proportion of amplitudes accumulated up to a given frequency in a measured parameter (e.g., pressure or force).

For blast experiments, we report the mean results of peak measurements for each scenario. For impact experiments, we report mean, sample standard deviation (STD), and coefficient of variation (COV) of peak measurements for each scenario.

Representative blast data for each transducer location from an unprotected and protected headform scenario can be seen in Fig. 6. Peak results and means for each protection scenario are tabulated in Table 3 for the incident blast and ground reflected waves, as measured by the external pressure transducers. Pressure versus time and FFT data from the ICP transducer measurements are shown in Fig. 7 for the three protection scenarios, with maximum and minimum values in Table 4. Similarly, force versus time and FFT data from the force transducer measurements are shown in Fig. 8 for the three protection scenarios, with maximum and minimum values tabulated in Table 5. Finally, normalized cumulative FFT data, which contain the cumulative sum of FFT magnitudes normalized by the total value accumulated up to the Nyquist frequency, are shown in Fig. 9.

Figure 7 shows representative external pressure field traces, which show near-instantaneous rises in pressure for the incident and ground reflective waves once they reach the headform, with an exponential decay in pressure after their arrival. For these experiments, the ground reflected wave arrives approximately 2 ms after the incident wave. Adding a helmet to the surrogate headform (scenario #2) reduced front incident shock magnitudes, shown in Table 3, from a mean of 224.5 kPa (scenario #1) to a mean of 194.2 kPa (attenuation of 13%); however, it increased the ground reflected shock magnitudes from a mean of 171.3 kPa to a mean of 231.4 kPa (amplification of 35%). It is important to note that this difference in trends noted in Table 3 for incident and ground reflected peak magnitudes is due to directionality of the arriving blast waves. The incident wave propagates directly toward the BIPED from the charge (at approximately the same height), resulting in purely frontal loading. Because the helmet brim in protected scenarios covers the external pressure transducers (Fig. 3(b)), the result is attenuation of the incident wave for helmeted trials relative to unhelmeted trials as the blast wave propagates through the helmet prior to reaching the external transducer. Conversely, the ground reflected wave propagates toward the BIPED at an upward angle, resulting in complex interactions between the helmet brim and scalp of the BIPED, yielding amplification of the ground reflected wave external pressures for helmeted trials relative to unhelmeted trials. Adding eye protection to the surrogate headform (scenario #3), in addition to a helmet, further attenuated the incident shock from a mean of 194.2–110.9 kPa (attenuation of 43%, relative to scenario #2), measured by the front external transducer. No systematic changes in overpressure measurements for the various protection scenarios were observed for the side and back external pressure transducers.

Measures from instrumentation at the front of the head suggest the addition of head protection affected ICP magnitudes and frequency content. From Table 4, peak frontal ICP measurements were reduced from a mean of 72.5 kPa in the unprotected trials (scenario #1) to means of 45.2 kPa and 37.0 kPa in scenarios #2, and 3, respectively. Adding a TBH helmet alone (scenario #2, relative to scenario #1) reduced mean frontal peak ICP measurements by 38%, while mean frontal external incident peak overpressure was reduced by only 13%. From Table 4, peak ICP signals for the side and back transducers did not convey obvious systematic alterations due to addition of protective headgear (despite a large reduction in the incident wave overpressure and front ICP). The FFT results in Fig. 7 demonstrate a shift in the ICP signal toward lower frequencies for the front and back ICP transducers, with reduction in amplitudes between frequencies from 1500 Hz to 2000 Hz. The side ICP data, however, did not convey the same shift in frequency. Side ICP data indicates greater spectral content in the 1000–1500 Hz range (protected relative to unprotected). The normalized cumulative distribution plots in Fig. 9 better illustrate the shift in frequency content, showing that more of the front ICP transducer signal is accumulated at lower frequencies, particularly below 2000 Hz, when head protection is used (scenarios #2 and 3) relative to an unprotected headform (scenario #1). This shift in spectral content toward lower frequencies is visualized in the normalized cumulative pressure plots as a leftward shift in the normalized cumulative pressure. For the exemplar plot in Fig. 9, approximately 60% of the front ICP transducer spectral energy is accumulated by 1000 Hz for scenarios #2 and 3, in comparison to approximately 47% for scenario #1. The side ICP data did not indicate change in normalized cumulative pressure for the various protection scenarios.

Relative to an unprotected head, using head protection under blast loading affects the force transducer measurements and their frequency content and the effects noted are largest for instrumentation located at the front of the head. From Table 5, front peak positive forces at the front are reduced from a mean of 8.52 N in unprotected blast trials (scenario #1) to means of 2.24 N and 1.67 N, for protected scenarios # 2 and 3, respectively (attenuation of 74–80%). In contrast, the side force increased from a mean of 3.56 N in unprotected trials (scenario #1) to means of 5.45 N and 5.36 N for protected scenarios #2 and 3, respectively (amplification of 50–53%). The FFT data in Fig. 8 also demonstrates a shift in spectral content toward lower frequencies for the front and back force transducers, with a particularly substantial attenuation in frequencies in the 1000–1500 Hz range for the front force transducer. Conversely, the side force transducers observe a shift toward more spectral content at higher frequencies, particularly in the 1000–1500 Hz range. Similar to the analysis of ICP signals, the normalized cumulative distribution plots in Fig. 9 better illustrate this phenomenon, showing that a greater amount of the force transducer signal is accumulated at lower frequencies when head protection is used (scenarios #2 and 3), relative to an unprotected headform (scenario #1). This shift in spectral content toward lower frequencies is shown as a leftward shift of the normalized cumulative force. For the exemplar plot in Fig. 9, between 40% and 50% of the front force transducer spectral content is accumulated below 1000 Hz for scenarios #2 and 3, in comparison to 30% for scenario #1. The normalized cumulative forces for the side transducers appear to accrue more in the 1000–1500 Hz range for the protected scenarios relative to the unprotected trials. Trends in minimum forces were overall relatively less evident than for maximum (positive) forces.

In Fig. 6, the reader can note increases in ICP (up to 4 kPa) and force (up to 0.5 N) that occur approximately 1 ms prior to the arrival of the incident blast. These increases have been noted in previous trials and are due to transmission of blast energy from the charge, downward to the ground, through the concrete at the test-site, and then into the head by way of the support post and Hybrid III neck. Because these ICP signals do not correspond to either the incident or ground reflected blasts, they are not considered in our analysis and discussion.

Representative impact data for each transducer location for both the unprotected and protected headform is shown in Figs. 10 and 11 for ICP and force measurements, respectively. Peak magnitudes with means of ICP and force are tabulated in Tables 6–8 for low, medium, and high-speed impacts, respectively. Plots comparing results (time series, FFT, and normalized cumulative measurements) for unprotected and protected BIPED scenarios, at each transducer location, are shown in Figs. 12 and 13 for ICP and force data, respectively.

During impact experiments, ICP and force peaks, measured nearest the impact site (front), are reduced by the addition of a helmet. In low-speed scenarios (Table 6), peak front ICP reduced from a mean of 32.5–15.5 kPa with the addition of a TBH helmet (52% reduction) and peak front force reduced from a mean of 6.83–1.18 N (83% reduction). In medium speed impacts (Table 7), peak front ICP reduced from a mean of 50.3–27.7 kPa (45% reduction) with the addition of a TBH, and peak front force reduced from a mean of 7.09 N to a mean of 1.16 N (84% reduction). In high-speed impacts (Table 8), peak front ICP reduced from a mean of 82 kPa to a mean of 46.3 kPa with the addition of a TBH (44% reduction), and peak front force reduced from a mean of 8.23 N to a mean of 2.57 N (69% reduction). Data from the back ICP transducer indicate decreases in pressure magnitudes with the addition of helmet. However, the side ICP transducer measurements do not indicate any systematic trend that could be attributed to the presence of head protection. In contrast, specific to force data (Tables 6–8), the back sensor did not exhibit trends of force reduction with the addition of a helmet. However, data from side located force transducers showed amplification of force with the addition of a TBH.

Intracranial pressure data for impact experiments do not exhibit shifts in spectral content with the addition of a helmet. The normalized cumulative pressures shown in Fig. 12 suggest that for pressure data, both protected and unprotected, over 70% of the accumulation resides between 0 Hz and 100 Hz. Force data did exhibit alteration in the accumulation of force amplitude between the cases of unprotected and protected BIPEDs (Fig. 13). Relative to an unprotected BIPED, the protected BIPED, for left and right located force transducer data, accumulates more amplitude at lesser frequencies. Data for front located force transducers suggest that the accumulation of amplitude with frequency is unaffected by the presence (or absence) of head protection. Force data associated with the back of the head suggest the unprotected BIPED accumulates amplitude at lower frequencies than the protected BIPED.

Using a repeatable simulant-based surrogate headform, instrumented with transducers capable of measuring pressure within the brain parenchyma and forces on the inner table of the skull, the results of this work suggest that head protection can be associated with overall reductions in pressures and forces measured near the site of loading. The surrogate, with novel instrumentation, and the data presented are the first that document forces, along with pressure within the brain, and the effects that adding head protection has on these measures in blast and blunt impact loading.

Free-field blast experiments using 5 kg C4 charges at a 5 m standoff produced repeatable blast loading overpressures measured on the outer surface of the BIPED. Coefficient of variation in measured external overpressure on the head exterior was on average 10% (however, it ranged up to 34% for back pressure measurement of incident shock in scenario 1). Relative to the alterations in peak pressures and peak forces in the cranial vault, associated with the addition of head protection, which in some cases were approximately 50% (pressure) and 80% (force), 10% variation is relatively small. Further, as shown in Table 3, while there was typically 10% variation (or less) in external pressure data within a given group of data for a given protection scenario, the variations in external pressures between protection scenarios were as high as 50% (reduced, relative to an unprotected BIPED). Consequently, we speculate that the alterations in force and pressure that are noted in the presence of head protection can be associated with the head protection, as opposed to nonrepeatability in loading between trials.

The overarching trend is that relative to an unprotected BIPED, head protection could be associated with attenuation of peak metrics. The majority of attenuation occurs for the transducer locations closest to the loading site, with peak forces at the front location attenuated by as much as 84% and peak pressures at the front location within the surrogate brain attenuated by as much as 52%. In contrast, back and side ICP transducer measurements were generally altered less and the magnitudes of pressures at these locations were lesser than the front. We speculate the small increase in forces observed on the side of the skull (e.g., Figs. 6 and 8) when adding head protection could indicate that adding head protection alters the mode of skull flexure [24] and alters the amount of lateral skull deformation under frontal blast loading, which results in increased lateral skull–brain interaction forces. It is also possible that in-plane loading of the transducers could confound measurements, and in future work we will assess the transducers for co-sensitivity to such in-plane loads.

The measurements from the BIPED also suggest that the addition of headgear alters the frequency content of forces and pressures experienced in the brain. For blast, Fig. 7 shows that higher frequencies are attenuated for the front ICP transducer, particularly in the 1500–2000 Hz range. Figure 8 also indicates attenuation for the front force transducer data, particularly in the 1000–1500 Hz range, but there is also attenuation from 800 Hz to 900 Hz. There appears to be a shift in spectral content with the addition of protective headgear, in which lower frequency peaks (observed below 600 Hz) are amplified while higher frequencies are attenuated (Figs. 7 and 8). The shift in spectral content is demonstrated in Fig. 9 where the normalized cumulative plots for the front ICP and force transducers accrue more amplitude at lower frequencies for the protected scenarios relative to the unprotected scenarios (observed as a leftward shift in the top two plots). Findings were mixed for instrumentation located at the side and back of the head; in some cases the protected head accrued amplitude at lesser frequencies, and in some cases greater, relative to the unprotected head. Figures 12 and 13 convey that relative to blast, alteration in frequency content due to head protection was relatively lesser for impact.

Intracranial pressure signals are significantly lower in comparison to the external overpressures (Fig. 7 shows data from various protection scenarios and transducer locations), indicating that attenuation occurs as the blast wave propagates through several materials (surrogate scalp, skull, membranes, cerebrospinal fluid (CSF), and then brain). Figure 6 shows that the most significant ICP fluctuations occur shortly after the arrival of the incident wave, and again approximately 2 ms later after the arrival of the ground reflection wave. Front ICP measurements, shown in Table 4, are approximately 20–40% of the front external pressure transducer, with a mean of 72.5 kPa for the front ICP transducer in an unprotected case (scenario #1), and means of 45.2 kPa and 37.0 kPa for protected cases (scenarios #2 and 3, respectively).

In blast experiments, both ICP and force transducers measured negative pressures and forces following positive values occurring earlier in the pressure–time and force–time data, despite a positive overpressure on the external head form. Arguably, this finding could be construed as counter-intuitive. In this study, ICP signals are observed to be as low as −40.5 kPa for blast and were not negative for impact. Inner table forces are observed to be as low as −6.2 N for blast and −2.2 N for impact. Negative pressures within the cranial vault have been measured in other work using cadavers, under blast loading, and computed in numerical simulations [42,43], in which it is suggested that skull flexure may be a mechanism through which negative pressures are created and cavitation in fluid occurs. While the thresholds of negative pressures that can lead to brain injury through cavitation in humans are not yet agreed upon, and indeed cavitation as a contributor to brain injury is arguably as yet unproven, the BIPED could be capable of quantifying negative valued pressure and force. While skull flexure and resonance have been studied in detail for the BIPED [24], further study is needed to ascertain whether or not the BIPED measurements are accurate in relation to cavitation mechanisms. The possible mechanisms for negative valued force measurements are expanded upon later in this discussion.

For the impact loading conditions we subject the BIPED to, the ICP data exhibit similarities to what has been published for experimental cadaveric and numerical simulations. A study on blunt impact to human cadaver heads reports coup and counter-coup pressures at velocities of 3–4 m/s [54], using a football helmet for head protection. They found mean peak coup pressures of 58±21.9 kPa (mean ± standard deviation) for a helmeted cadaver head, and 68.1±47.6 kPa for mean peak coup pressures for an unhelmeted cadaver. Impact pressure pulse widths varied significantly but were typically between 5 and 35 ms duration. ICP and force versus time data from the BIPED exhibit similar pulse widths (approximately 10–20 ms for most trials) to the average cadaveric results. Peak frontal ICPs of 82.0 kPa without a helmet and 46.3 kPa with a combat helmet were observed within the BIPED for the high-speed impact scenario (3.3–3.4 m/s), which are within the pressure ranges observed during the previous cadaveric study, for a similar impact velocity.

The BIPED surrogate model used in this study demonstrates repeatability in both impact and blast measurements for the front located instrumentation, in both unprotected and protected scenarios. For example, in high-speed impact experiments, peak front force signals and peak front ICP signals had COV less than 7% in both protected (n = 5) and unprotected scenarios (n = 5). Impact pulse widths (calculated based on the initial impact signal width at 20% of the peak value, reported in a cadaver study by Hardy et al. [54]) are also significantly more repeatable in the BIPED compared to experiments with post mortem human subjects [54], with COV for ICP and force pulse widths less than 12% for each impact scenario.

We noted important differences in the forces measured at the inner table when comparing impact results to blast results. During impact, force transducers at the side and back of the head measure negative values, while at the front of the head measure positive values. The time-duration of these forces (nominally 20 ms) are approximately the same when comparing front located sensors to all other locations. The negative valued forces could be interpreted to suggest that the separation between the inner table and simulant brain increases as the impact event occurs causing net reduction in the amount of force that the brain exerts on the skull. The separation could be due to either motion of the brain relative to the skull (while the skull geometry is unchanged), or flexure of the skull toward or away from the brain, or a combination of both. The positive increasing forces at the front could be due to the brain compressing the skull due to relative motion, or skull flexure leading the skull to press against the brain, or both. In blast, the force data exhibits high-frequency sign reversal in force magnitude, at timescales lesser than 3 ms, and these sign reversals appear at all locations. It is arguably plausible that skull-brain motion and or skull flexure be responsible for the sign reversals measured. Detailed examinations in previous studies suggest that for blast, these sign reversals correspond to skull flexure mechanics [24]. In future work, we will attempt to ascertain whether the mechanics-based explanations offered are valid, and if it is, the relative contributions of skull flexure or relative motion between the skull-brain in the force versus time data to ascertain which mechanism dominates. Strain sensors, that are mounted to the skull and colocated with force transducers, could transduct skull flexure and allow us to ascertain the proportional contribution that skull flexure has on energy transfer in impact and blast loading to the BIPED.

The findings in this study, as with any study that involves a surrogate model, are subject to limitations and may not be indicative of how a human head would respond under the same loading conditions. While the BIPED has been shown to measure similar ICP magnitudes and fluctuation durations to a cadaver headform under a specific set of blast loading conditions [69], this does not imply that the BIPED will respond equally to a human head in other loading scenarios such as the free-field blast conditions and the impact loading conditions we tested in this study. In addition, there is no literature in which skull-to-brain contact forces have been measured in the living or in post mortem subjects, and therefore we acknowledge that our novel techniques leading to force measurements on the inner table cannot, at this time, be compared to the work of others. Further, we must, therefore, acknowledge that the force measurements in the present work should be verified ex vivo so that a comparative dataset can be produced that allows BIPED measures of force to be compared to those in post mortem human subjects. In addition, the hybrid III neck arguably may not be an appropriate model of the human neck for blast. It is often criticized as possessing too great a mechanical stiffness and researchers have speculated the stiffness may lead to head kinematics that are unrealistic relative to the living human, specifically in blunt impact [77,78]. Because the hybrid III neck has not been characterized or designed for applications in blast, we acknowledge that it could be a poor model for application in blast and further that the findings of our study could be altered if neck mechanics are altered.

For tertiary blast injury, we focused on head impact, simulating a scenario where a soldier falls upon a soft surface (modeled using artificial turf). We acknowledge that there are many plausible scenarios associated with tertiary effects, including factors associated with impact orientation and impact surfaces, which were not included in this study. For both blast and impact experiments in this study, the loading into the head was always frontal. Further experiments would be needed to assess whether head protection leads to reductions in forces and pressures in other loading directions. Future work could also involve testing secondary blast injury due to ballistic/projectile threats from debris/shrapnel [23].

The reader should also note that, for blast experiments specifically, that the number of experiments performed led to sample sizes that would be considered too small to perform statistical tests and therefore draw statistical inferences. We highlight this limitation to reinforce that the findings in this study should be interpreted as trends to be studied with further experimental work. In addition, in part due to the small number of repetitions for blast experiments, we have not tabulated COV as we felt that the sample sizes may not capture true variance.

In this work, we quantify mechanical measures of forces on the inner table and pressure within the brain parenchyma of a simulant-based surrogate headform subjected to head-level face-on blast exposure and high forehead impact loading. Our results suggest that adding head protection can alter the measured forces and pressures relative to the case when the surrogate model is unprotected. Head protection had the greatest influence on measurements nearest to the site of blast/impact loading, with front forces attenuated by as much as 80% under blast loading and 84% under impact loading, and front ICP signals attenuated by as much as 49% under blast loading and 52% under impact loading. Additionally, results for forces and pressures during blast (measured at the head front) for a helmeted head demonstrate propensity toward more amplitude accrual at lower frequencies, relative to an unprotected head. It is premature to ascribe definitive statements regarding mitigation of brain injury risk based on our mechanical measures and therefore the measurements described in this work should not be construed to indicate trends or efficacies/deficiencies in performance of protective headgear. However, it is possible that with further biomechanical research that elucidates how mechanics relates to brain injury likelihood in blast and impact, that measurements like those presented here could indicate whether or not protective headgear offers net protective effects to military personnel exposed to blasts and impacts.

We gratefully acknowledge technical support and collaborative agreements from the Weapons Effects and Protection Section Personnel at the Valcartier Research Centre of Defense Research and Development Canada and also Defense Scientists at Defense Research and Development Canada's Suffield Research Center. This work is financially supported by the U.S. Department of Defence Army Research Labs through the U.S. Army International Technology Center - Canada Offices. We gratefully acknowledge support from the Natural Sciences and Engineering Research Council (Canada) Discovery Grants and Research Tools and Instruments Grants, as well as the Faculty of Engineering and Department of Mechanical Engineering at the University of Alberta. In addition, we would like to recognize the support of the Canada Research Chairs Program.

Research was sponsored by the Army Research Laboratory and was accomplished under Cooperative Agreement Number W911NF-16-2-0083. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Army Research Laboratory or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein.

• Army Research Laboratories (W911NF-16-2-0083; Funder ID: 10.13039/100006754).

• Canada Foundation for Innovation (John R Evans Leaders Fund - High speed camera system for imaging the mechanics of Impact) (Funder ID: 10.13039/501100001805).

• Natural Sciences and Engineering Research Council (Canada) (Grant No. RGPIN 018-04253; Funder ID: 10.13039/501100000038).