The French Adrian helmet performs better than modern designs


Biomedical engineers from Duke University have demonstrated that, despite significant advancements in protection from ballistics and blunt impacts, modern military helmets are no better at protecting the brain from shock waves created by nearby blasts than their World War I counterparts.

And one model in particular, the French Adrian helmet, actually performed better than modern designs in protecting from overhead blasts.

The research could help improve the blast protection of future helmets through choosing different materials, layering multiple materials of different acoustic impedance, or altering their geometry.

The results appeared online on February 13 in the journal PLOS ONE.

“While we found that all helmets provided a substantial amount of protection against blast, we were surprised to find that the 100-year-old helmets performed just as well as modern ones,” said Joost Op ‘t Eynde, a biomedical engineering PhD student at Duke and first author of the study.

“Indeed, some historical helmets performed better in some respects.”

Researchers have only recently begun to study the brain damage a shock wave can cause on its own — and for good reason. Helmets were originally designed to protect from penetrating objects like bullets and shrapnel, and blast waves will kill through pulmonary trauma long before they cause even minor brain damage.

With the advent of body armor, however, soldiers’ lungs are much more protected from such blasts than they used to be.

This has caused the incidence of pulmonary trauma following a blast to drop far below that of brain or spine injuries in modern military conflicts, despite the difference in blast tolerance.

While there have been studies that suggest modern helmets provide a degree of protection from shock waves, no currently deployed helmet has been specifically designed for blast protection.

And because soldiers today experiencing shock waves while wearing body armor aren’t all that different from soldiers 100 years ago experiencing shock waves while in the trenches, Op ‘t Eynde decided to see if those old designs offered any lessons to be learned.

“This study is, to the best of our knowledge, the first to assess the protective capabilities of these historical combat helmets against blasts,” said Op ‘t Eynde.

Working with Cameron “Dale” Bass, associate research professor of biomedical engineering at Duke, Op ‘t Eynde created a system to test the performance of World War I helmets from the United Kingdom/United States (Brodie), France (Adrian), Germany (Stahlhelm) and a current United States combat variant (Advanced Combat Helmet).

The researchers took turns placing different helmets on a dummy’s head outfitted with pressure sensors at various locations.

They then placed the head directly underneath a shock tube, which was pressurized with helium until a membrane wall burst, releasing the gas in a shock wave.

The helmets were tested with shock waves of varying strength, each corresponding to a different type of German artillery shell exploding from a distance of one to five meters away.

The amount of pressure experienced at the crown of the head was then compared to brain injury risk charts created in previous studies.

While all helmets provided a five-to-tenfold reduction in risk for moderate brain bleeding, the risk for someone wearing a circa-1915 French “Adrian” helmet was less than for any of the other helmets tested, including the modern advanced combat helmet.

“The result is intriguing because the French helmet was manufactured using similar materials as its German and British counterparts, and even had a thinner wall,” said Op ‘t Eynde.

“The main difference is that the French helmet had a crest on top of its crown. While it was designed to deflect shrapnel, this feature might also be deflecting shock waves.”

This shows the WW1 and a modern helmet getting shock tested

A French helmet from World War I and a modern helmet sit beneath a shock tube to test how well they protect the dummies underneath from a shock wave.

The ridge down the center of the French helmet was designed for deflecting shrapnel, but may well have also helped deflect the shock wave, allowing the helmet to outperform even modern combat helmets. Image is adapted from Duke University news release.

It also might be that, because the pressure sensor was mounted directly under the crest, the crest provided an additional first layer for reflecting the shock wave.

And the French helmet did not show the same advantage in pressure sensors at any other location. For locations such as the ears, performance seemed to be dictated by the width of the helmet’s brim and just how much of the head it actually covered.

As for the modern helmet, Op ‘t Eynde theorizes that its layered structure might be important in its performance.

Fig 1. Coronal CT sections of tested helmets.(a) Adrian helmet, (b) Brodie helmet, (c) Stahlhelm, (d) Advanced Combat Helmet (ACH). The Adrian helmet (a) is the thinnest steel followed by the Brodie (b) and the Stahlhelm (c). The ACH (d) is made with a fiber composite material.    

Because a shock wave is reflected every time it encounters a new material with a different acoustic impedance, the layered structure of the modern helmet might contribute to its blast protection.

But no matter which helmet was tested, the results clearly indicated that helmets might play an especially important role in protecting against mild blast-induced brain trauma.

According to the researchers, this finding alone shows the importance of continuing this type of research to design helmets that can better absorb shock waves from nearby overhead explosions.

“The difference a simple crest or a wider brim can make in blast protection, shows just how important this line of research could be,” said Op ‘t Eynde, who initially came to Duke on a scholarship from the Belgian American Educational Foundation, which was established with funds from American relief efforts in Belgium during World War I.

“With all of the modern materials and manufacturing capabilities we possess today, we should be able to make improvements in helmet design that protects from blast waves better than helmets today or 100 years ago.”


Three authentic historical WWI infantry combat helmets including the original lining, were acquired for blast testing: an M15 (1915 model) Adrian Helmet used by the French Army (denoted FRC), an M1916 Stahlhelm used by the Imperial German Army (denoted GER), and an M1917 Brodie Helmet used by the U.S. Army (based on the M1915 British design and denoted AMR).

The M1917 Brodie Helmet was manufactured by the Columbian Enameling and Stamping Company (Terre Haute, IN, USA). The Advanced Combat Helmet, the current combat helmet used by the U.S. Army, was included (size large, denoted ACH) for comparison to current combat helmets. A ‘no helmet’ bare head case was used as a control (denoted BAR).

The three WWI helmets are made of formed steel, and the Advanced Combat Helmet (ACH) has a fiber composite construction. The average wall thickness of each helmet was measured using electronic calipers (EC799, L. S. Starrett Company; Athol, MA, USA).

The projected area for the top view of each of the helmets was calculated using ImageJ (NIH; Bethesda, MD, USA). Weight, wall thickness, and projected area of each of the helmets, and abbreviations used in this manuscript are described in Table 1. High resolution X-ray computed tomography images(Nikon XTH 225 ST, Nikon Inc.; Minato, Tokyo, Japan) were made of the historical helmets and coronal sections are displayed in Fig 1.

Blast setup

Helmets were mounted on a Hybrid III® 50th percentile male dummy head (Humanetics; Farmington Hills, MI, USA) and affixed to the Hybrid III neck. Each helmet was secured around the chin and back of the dummy head (Fig 2) to prevent extraneous helmet motion during testing.

Original buckles and leather straps were not used due to the degraded conditions that would not withstand the blast scenarios.

For the ACH, original helmet straps were used. The ACH fit properly on the dummy head, covering the forehead while leaving one to two centimeters space above the eyes as described in the ACH operator’s manual.

The historical helmets all fit on top of the head, with the head held in the internal suspension without the crown of the head touching the helmet. Each helmet had both the external steel components and internal textile/leather components intact.

The dummy head was faced downwards, and the center of the head was aligned with the open end of a cylindrical blast tube (schematic in Fig 3). This orientation and blast exposure simulate an overhead blast scenario, as would have been common in trench warfare due to artillery shells exploding above trenches.

Fig 2. Helmets on Hybrid III head in test setup.(a) Brodie helmet, (b) Adrian helmet, (c) Stahlhelm, (d) Advanced Combat Helmet, (e) No helmet.  

The top of the helmet was aligned with the end of the blast tube to minimize standoff distance. The blast tube has a diameter of 305 mm and consists of a driver section (305 mm length), where helium gas is compressed, and a driven section (3.05 m length).

The driver and driven section are separated by a diaphragm consisting of a number of polyethylene terephthalate (PET) membranes. High pressure helium is released into the driver section until the PET diaphragm bursts, allowing a shock wave to travel down the driven section of the blast tube.

The 10:1 driven length to driver length ratio allows the shockwave to develop a uniform shock front, with approximate equal pressure across the tube section [31]. In a previous study [31], it was shown that testing outside the blast tube is appropriate as long as standoff distance is minimized.

The helmets were exposed to shock waves at three separate blast intensities by varying the thickness of the bursting diaphragm with PET membranes: two membranes of 0.254 mm thickness (total thickness: 0.508 mm), nine (2.286 mm), and twelve (3.048 mm).

These choices for membrane thickness and resulting shock intensity were made to represent historical blast exposure (see blast simulation) and approximate blast levels corresponding to 50% risk for respectively mild meningeal bleeding, moderate meningeal bleeding, and severe meningeal bleeding based on bare head ferret brain blast data from Rafaels et al. [24]. In total, forty-six blast tests were performed for this study, detailed in Table 2.

All two-membrane tests were performed first in the order ACH-FRC-AMR-GER-BAR, followed by all nine-membrane tests in the order BAR-GER-AMR-FRC-ACH, and finally all twelve-membrane tests in the order ACH-AMR-FRC-GER-BAR. For each helmet all tests at a specific number of membranes were performed consecutively.


In total, forty-six blast tests were performed for the five helmets over three blast conditions described in Table 2. After observing deformations that may affect the structural integrity in the Brodie and Adrian helmet (Fig 7) at the highest tested peak pressures, it was decided to keep the number of 3.048 mm PET tests at two for the historical helmets. For all other exposures, helmets experienced minimal deformation and no evidence of degradation by repeated blast exposure was observed. Blunt impact of helmet deformation on the head was not assessed in this study.

Fig 7. High speed frames of blast impact on the head.Frames from high speed video recording of high intensity blast tests immediately after blast wave impingement on the (a) Brodie helmet and (b) Adrian helmet.    show more

Peak tube pressure, helmet type, and their interaction were each found to have a statistically significant effect on the crown pressure for each measurement location (p<0.001). Statistical significance of the interaction term between helmet type and tube pressure justifies the use of different regression slopes for the different helmet types.

At the crown of the head, the interaction term differed significantly between the bare head and all helmets (p<0.0001), with higher crown pressures for the bare head. There was also a significant difference in slope between the ACH and the French Adrian helmet (p<0.0001). When removing the interaction term, the Adrian Helmet results showed a significant difference in pressure compared to the British/American Brodie helmet and the German Stahlhelm (p<0.01).

The Adrian helmet resulted in lower crown pressures than all other cases. The ACH, Stahlhelm (GER), and Brodie helmet (AMR) were not found to be significantly different from each other (p>0.05).

The results of the general linear model for the crown measurement location are shown in Fig 8. Some of the regression lines for the helmet blast results do not pass through the origin, suggesting that the peak pressure attenuation provided by helmets might be nonlinear at lower blast pressures.

Besides lower peak pressures, the crown pressure traces also showed a more gradual loading rather than a near instantaneous shock front when assessing the helmeted case compared to the bare head (Fig 5).

Video analysis of the blast event showed that the delay in pressure rises seen in the figure corresponds to the helmet moving in the suspension towards the head, and the peak pressure time roughly corresponds to the maximum compression of the helmet suspension. However, because of the short durations (< 2 ms) of these pressure peaks, the use of a blast injury criterion was deemed appropriate.

The blast tests carried out at different amplitudes were found to be in the 50% risk range for mild, moderate and severe meningeal bleeding for crown pressure on the bare head (Fig 9) based on the scaled ferret risk curves (section 2.5) [24]. Wearing a helmet was associated with a decrease in bleeding risk.

This shows that the performed tests simulate realistic exposures where wearing a helmet might change physiological outcomes in the brain. In Fig 10, the 50% moderate meningeal bleeding case for the bare head is compared to the helmet results at that level. For the same blast conditions, risk of moderate bleeding is lower than 10% in all helmets, and close to 1% for the Adrian helmet.

At the forehead measurement location (Fig 11a), there was a significant difference in slopes between the bare head and all helmets (p<0.0001), and the Stahlhelm was different from the ACH and Brodie helmet (p<0.05) with higher pressures. No additional differences were found when the interaction term was removed (p>0.05).

For the back of the head (Fig 11b), there was again a significant difference in slopes between the bare head and all helmets (p<0.0001). When removing the interaction term to consider equal slopes, the Stahlhelm was found to differ from the Brodie helmet and ACH (p<0.01), and the Adrian helmet also differed from the Brodie helmet and ACH (p<0.005).

For pressure measured at the left eye (Fig 11c), the bare head and the ACH both differ in slopes compared all other helmets (p<0.0001) with higher pressures. The Adrian helmet slope differs significantly from the Brodie helmet (p<0.001). With equal slopes, the bare head has a significantly higher pressure than the ACH (p<0.0001).

At the right ear measurement location (Fig 11d), the bare head had a steeper slope than all helmets (p<0.0001). Comparing with equal slopes, the Adrian helmet had a higher pressure than the Brodie helmet and ACH (p<0.01), and the Stahlhelm (p<0.05).

The ear pressures for the bare head condition exceeded 50% major rupture risk [33] for all tested severity levels (Fig 12). Rupture risk was reduced in all helmeted conditions, with less than 50% risk for a minor rupture at the low severity levels for all helmets except the Adrian helmet and 50% moderate to 50% major rupture risk at the medium and high severity levels.


Blast exposure to the bare head was more severe than any helmeted test for every blast intensity and at every measurement location. The bare head experienced three to five times higher peak pressures (Fig 8) at the crown of the head (at similar positive phase durations), which corresponds to higher risk of meningeal bleeding and other potential brain injuries [24].

Helmets provided more shock wave attenuation at lower pressure levels than at higher pressure levels (Fig 8), suggesting that helmets might play an especially important role in protection against mild primary blast induced brain trauma.

The effect of wearing a helmet, especially for short positive phase durations (0.5–5 ms), is a significant reduction in risk of blast brain injury at the crown of the head for overhead blast scenarios.

In other orientations, blast wave measurements are complicated by the difference between reflective (measured with pressure gauges oriented parallel to the direction of the blast) and incident (measured with pressure gauges oriented perpendicular to the direction of the blast) pressures, leading to conflicted reports of helmets possibly increasing the risk of primary blast injury [3642].

This risk has to be carefully evaluated because reflected pressure measurements can be two to eight times greater than incident pressure measurements for the same blast scenario [32].

An interesting result from these experiments is the blast protective effect provided by the French Adrian helmet, which had a lower crown pressure than all other helmets, despite being manufactured using similar materials as the Stahlhelm and Brodie Helmet, with a thinner helmet wall (Table 1).

This result might stem from the deflector crest along the midline of the helmet (Fig 1a). Specifically added with overhead shrapnel in mind [43], this feature of the helmet could deflect the shock wave off to the side of the head, rather than allow shockwave impingement onto a more planar surface seen in the other helmets.

The crest also provides an added first layer for shock wave reflection before reflecting a second time off the helmet itself. The crown pressure sensor used in the measurements was located under the deflector crest and may have experienced a decreased peak pressure because of this. Further studies are needed to see if surface geometry manipulation or helmet attachments may augment the protective capabilities of helmets against blast exposure.

Peak pressures measured in locations other than the crown of the head were much lower because of measurement at an orientation incident to the blast wave and being partly or completely covered by the helmets. In these locations, the Adrian helmet did not provide the same protective advantage seen at the crown.

Pressure attenuation was seemingly determined by the width of the brim and/or coverage of the helmet (Fig 2). At the ear, the small brim and limited coverage of the Adrian helmet resulted in higher pressures than the other helmets (Fig 11d), with a corresponding increased risk in eardrum damage (Fig 12).

The ACH, without a brim as seen in the historical helmets, had increased pressures at the eye (Fig 11c) but provided similar protection at the other measurement locations.

While ballistic protection provided by helmets has increased significantly since WWI and saved many lives [39], the results found here suggest that the ACH did not perform quantitatively or qualitatively better than the historical helmets, and performed worse than the Adrian helmet for overhead primary blast at the crown of the head.

On the other hand, while ballistic protection has been an active focus in combat helmets design, protection from primary blast has not been an important design element [39], and the level of protection from primary blast from all of the helmets tested is large compared with the bare head. One of the reasons for this is that the mechanism for blast protection was poorly understood for the first sixty years following WWI.

While the exact injury mechanism for primary blast is still unknown, the scientific community (cf. Cooper, 1991) [44] identified acoustic impedance as one important protection mechanism against blast waves.

The acoustic impedance protection mechanism against blast trauma is different than against ballistic trauma. An ideal protection against ballistic impacts can locally absorb high energy impacts without failure or excessive deformation by distributing the energy through the material [45].

Desirable materials have high strength, high modulus, and a high local speed of sound. Protection from primary blast waves can be obtained by attenuating the blast wave using an acoustic impedance mismatch at an interface the wave is travelling through.

An increased difference in acoustic impedance causes a higher proportion of the blast wave to be reflected, rather than penetrate into the body where it causes local stresses and tissue damage [44]. The reflection coefficient R can be calculated from Eq 2.


In Eq 2, Zhelmet is the acoustic impedance of the helmet and Zair is the acoustic impedance of the air. Acoustic impedance Z is calculated as the product of speed of sound in the material and density of the material. Ideal materials have a high local speed of sound and a high density.

Steel has a greater acoustic impedance (~38⋅106 Pa⋅s/m3 for hardened manganese steel [46], used in WWI helmets) than composite fibers (~12⋅106 Pa⋅s/m3 for Kevlar® 129, used in ACH [47]), but since both impedances are orders of magnitude higher than air (~440 Pa⋅s/m3), reflection will be relatively similar (R = 0.999977 for steel and R = 0.999927 for Kevlar® 129).

This explains the similar results for the ACH, Brodie helmet, and Stahlhelm. Many helmet and body armor materials have properties that are desirable for both ballistic and blast trauma. Because a shock wave reflection occurs at every interface where there is an acoustic impedance mismatch, primary blast protection can be improved by using multi-layered configurations of high and low acoustic impedance, with each layer reflecting a proportion of the penetrating wave.

Not every layer of material will be beneficial to blast wave protection, and if a material has an acoustic impedance in between two neighboring materials, it will enhance blast wave penetration. The layered structure of the ACH might contribute to its blast protection, but future studies are needed to evaluate the effect of a layered structure.

Helmet wall thickness improves ballistic protection by providing higher strength and energy absorption, but it doesn’t affect blast protection much since reflection only occurs at interfaces. While the Adrian helmet provided superior blast protection at the crown of the head for overhead blast in this study, Dean [1] noted that the ballistic protection it provided was less than both the Brodie helmet and Stahlhelm.

One of the limitations of this study is that only an overhead blast scenario was examined. While this would be an accurate approximation of blasts in trench warfare as in WWI or air bombings of soldiers in the field during major unit action, it would not be as applicable to other cases such as improvised explosive devices (IEDs) used as roadside bombs, a significant cause of injury and death in conflicts in Iraq and Afghanistan [48].

The current study evaluated primary blast protection without considering reflective surfaces. In combat scenarios, reflection of a blast wave off of a surface can change outcomes considerably [49], such as when a soldier lies on the ground with the crown of the head towards the blast, or is confined within a trench.

Another limitation is that the historical helmets tested are over one hundred years old, and their material properties might not be the same as they were originally. While properties of steel are relatively stable, the helmet linings may have degraded. However, there is no guarantee that replicas would be identical copies of the original either, so this study stays as true to the original helmets as possible.

Finally, this study did not include the potential from blunt neurotrauma from impacts of the helmet on the head following acceleration from the transiting shock overpressure. This effect may be large with blasts that had larger positive phase duration and larger impulse than for the shells considered in this study.

Duke University


Please enter your comment!
Please enter your name here

Questo sito usa Akismet per ridurre lo spam. Scopri come i tuoi dati vengono elaborati.