Surrey
A GREENISH-YELLOW disembodied head, perched on a neck 30 centimetres high,
takes pride of place in Sandra Bell鈥檚 laboratory. Soon it will be getting
kicked, punched, bounced on the floor, hit with a hammer or placed inside a
speeding car about to collide with a wall. But for now it stares mutely into
space, silently contemplating its fate.
Bell鈥檚 work might sound ghoulish, but in fact the disembodied head is a
dummy鈥攑robably one of the most realistic and sensitive dummies in the
world. It has been designed to find out how and why humans suffer brain damage
by recording the effects of stress and shock waves generated during sudden
impacts. The fake head is made from new materials which mimic the texture and
physical properties of everything from brains and skin to skull bone. And it is
chock-full of piezoelectric sensors and devices for measuring motion.
Advertisement
Bell, a physicist at Britain鈥檚 Defence Evaluation Research Agency (DERA), is
leading the research project involving the dummy, launched this week to attract
commercial collaborators. The dummy head, known as 鈥淒ERAMan鈥, will suffer car
crashes, boxing blows, sporting collisions and even being fired from ejector
seats to discover the most damaging stress waves produced by each. Bell hopes
this will help designers make safer vehicles and sports equipment, reducing the
risk of brain damage.
Around 100 000 people die each year worldwide from head injuries, and as many
as 90 000 more are left disabled. But no one yet understands what happens in the
brain when it is involved in something like a car crash. Conventional crash
dummies are simply designed to have the same shape and mass as an average human
body, with joints in the right places to simulate the physical movements of a
body during an impact. But although they have led to radically improved car
safety, these dummies have no brains.
鈥淲hen it comes to predicting head injuries, the dummies currently used are
extremely crude,鈥 says Tony Payne, a consultant at the Motor Industry Research
Association in Nuneaton. Heads are simply aluminium shells containing an
accelerometer that measures up and down, forwards and backwards and side-to-side
forces felt by the head. From these measurements a 鈥渉ead injury criteria鈥 (HIC)
number is worked out, and the rough assumption is that a HIC greater than 1000
is fatal. But no one even knows whether the HIC is a valid assessment.
Shock waves
Brain injuries are not simply caused by a physical impact producing, say, a
skull fracture: severe brain injuries occur without the skull being damaged,
says Fausto Iannotti, a researcher at the Department of Clinical Neurological
Sciences at the University of Southampton. 鈥淎ny sudden acceleration of the head
causes the brain to move within the skull,鈥 he says. The energy transferred to
the head passes through the brain as two types of stress wave鈥攕hear waves
and compression waves. Iannotti believes that each can cause a distinct type of
brain damage, and Bell is aiming to detect both in her dummy.
Shear waves can be seen clearly in jelly when you wobble it on a plate, says
Christopher Morphey, of the Institute of Sound and Vibration Research, also at
Southampton. 鈥淚f you suddenly rotate the plate, the jelly at the bottom will
move immediately,鈥 says Morphey. 鈥淏ut the top of the jelly will begin to move a
fraction of a second later. The jelly is twisted, creating shear stresses in the
structure.鈥 If the rotation is fast enough, the shear stresses can rip the jelly
apart.
Now read 鈥渂rain鈥 for 鈥渏elly鈥. 鈥淪hear waves pass through the brain at between
1 and 10 metres per second,鈥 says Philippe Young, a dynamics engineer at the
University of Exeter. A shear wave can take 20 milliseconds to pass from the
surface to the centre of the brain鈥攔oughly the duration of an impact. So
the centre of the brain is just starting to move as the outside is stopping.
Iannotti thinks that the resulting twist is an important cause of brain injury.
鈥淎t the junction of the spinal cord and the brain, such twisting may fracture
the tissue,鈥 he says, 鈥渁nd in the bulk of the brain, axons lying in the
direction of the deformation might be stretched and damaged.鈥
Boxers鈥 brains often show a form of damage known as diffuse neuronal injury,
in which axons and neurons are damaged throughout the brain. Shear waves could
be responsible鈥攇lancing blows to the head could cause the whole brain to
twist sharply and fracture in many areas.
Compression waves鈥攕hock waves from an impact which travel at around
1500 metres per second and cause the brain to squash or expand like an
accordion鈥攃an damage the brain in a different way. Bell believes that
compression waves could be to blame for 鈥渃ontra coup鈥 injuries, in which brain
damage occurs on the opposite side of the head to the impact. The idea is that
the shock wave passes through the brain, hits the skull at the other side, and
most of the energy is reflected back. As the wave is reflected, it causes an
extreme drop in the internal pressure of the brain for a few milliseconds,
enough to boil blood and rupture vessels.
Skull reflections
鈥淭his is what engineers call cavitation,鈥 says Young, who is modelling the
effects of shock waves in a simple oval-shaped head. Every oval has a particular
frequency, or resonant frequency, at which it vibrates, and Young鈥檚 model shows
that cavitation is worse when the frequency of the shock wave produced by an
impact matches the oval鈥檚 resonant frequency. Reductions in pressure could even
occur on both sides of the skull as the wave reflects backwards and forwards.
鈥淭his suggests that cavitation may be the cause of [brain] injury鈥攏ot only
at the contra-coup site, but also directly under the site of injury,鈥 he
says.
Multiple reflections of shock waves from different parts of the skull could
also create problems. The waves could combine to give regions of higher or lower
pressure, creating an interference pattern in the brain, just as music is louder
or softer at different places in a concert hall depending on the interaction of
the sound waves. Bell thinks that the regions of high pressure may cause the
widespread damage seen in diffuse neuronal injury. But Morphey is sceptical.
鈥淥ur models indicate that there are no dramatic high and low spots caused by
interference patterns,鈥 he says.
DERAMan should settle the debate. Bell reckons that her dummy will identify
the frequencies of the shock waves, and let researchers map all the pressure
changes within the brain during an impact. The head contains a full brain of
sensors to detect both compression and shear waves, as well as forces of motion.
It has 40 pressure sensors made from polyvinyledene difluoride, a piezoelectric
polymer that generates a small electric charge when squeezed, which are buried
in the middle of the brain sandwiched between the two cerebral hemispheres.
Another 45 ceramic piezoelectric sensors line the inner surface of the
skull.
Add to these signals the readings from two accelerometers and a
three-dimensional force gauge, and DERAMan has 90 channels of information which
are fed into a computer for real-time analysis. 鈥淭his allows us to build a
three-dimensional understanding of the way waves propagate within the artificial
brain,鈥 says Bell.
Tailor-made flesh
But obviously all those data would be worthless if the artificial brain
didn鈥檛 accurately imitate the response of natural tissues to impacts and the
stress waves produced. To achieve this, the materials must mimic the brain in
the way they reflect, transmit and dissipate energy鈥攁 property known as
impedance. David Townend, an acoustics expert at DERA, can produce materials
with specific impedances using a special polymer. By altering the composition of
his polyurethane system, Townend can tailor-make materials with the dynamic
properties of human flesh.
DERAMan has a custom-made polyurethane outer skin, a hard plastic skull and a
brain made from another polyurethane with a different composition. 鈥淭he
materials are probably not optimal at the moment,鈥 says Townend, 鈥渂ut they are
considerably better than anything existing up until now.鈥
Townend鈥檚 polyurethanes are made from three basic components. 鈥淚t鈥檚 a bit
like steel scaffolding,鈥 he says. Long rod-like polyols make up the main
structure, triol molecules form cross-links, and isocyanate groups act as chain
extenders, linking the structure together. By mixing the components in different
proportions, Townend can produce polymers with a wide range of
impedances鈥攊ncluding those of the requisite human tissues.
The technology behind the new materials comes from research into making
submarines invisible to sonar, explains Townend. Submarines are coated with a
polymer of a specific impedance that absorbs sonar waves and disperses the
energy rather than reflecting them.
Impedance is determined by the rate at which polymer chains can rotate, known
as the molecular relaxation frequency. This frequency depends on how rigid or
stiff Townend makes the polymer鈥攖he more triols and isocyanates, the
stiffer the material and the slower the rotation.
When the molecular relaxation frequency matches that of a stress wave, energy
is absorbed causing the polymer chains to rotate. Similarly in the human brain,
the waves with frequencies that pass their energy into the tissues are the
culprits in brain damage.
To confirm that DERAMan鈥檚 materials do indeed mimic the human brain during an
impact, Bell鈥檚 team will be comparing its findings with clinical data from
people with fatal head injuries. This will involve estimating the site,
direction and likely force of an impact and correlating this with the damage
seen in MRI scans, X-rays and post-mortem examinations.
If they match up, then reconstructing the same sort of accident using DERAMan
will help work out which injuries are due to which kind of wave. For example,
DERAMan may be attached to a standard dummy body for car crash testing. The
stress waves generated by any impact on the head could be isolated, and
manufacturers might then redesign the components of the car interior鈥攕uch
as dashboard, windscreen surround, seating鈥攖o cushion the blow. 鈥淚f you
don鈥檛 consider the problem fully you could end up making the situation
considerably worse [with a new design],鈥 warns Townend. The ideal solution would
be to channel the stress waves from an impact away from the person, into the
dashboard and the rest of the car. 鈥淎bove all you don鈥檛 want to trap the energy
in the person鈥檚 head,鈥 he says.
Bell also hopes to get a professional boxer to punch DERAMan, in an attempt
to pinpoint the cause of brain damage in boxers. 鈥淭he frequencies of [stress]
waves that come from being hit by a boxing glove are similar to those of a blast
wave from a small charge of high explosive,鈥 says Townend. Analysing DERAMan鈥檚
response to a typical punch might help to work out which frequencies are most
damaging and then manufacture boxing gloves from special materials designed to
control the frequencies of stress waves produced and the way that energy passes
from glove to head.
Adrian Whiteson, chief medical adviser to the British Boxing Board of
Control, is keenly anticipating Bell鈥檚 results. 鈥淎nyone in a sport that involves
trauma to the head would be very interested,鈥 he says.
Sports don鈥檛 come much more traumatic than American Football. Every piece of
action involves players in some kind of impact, and to prevent physical injury
they are extensively padded. But experiments with DERAMan could improve the
padding, and particularly the helmets, to take account of collisions with
opponents. 鈥淗ead injuries are a factor in American Football,鈥 says Dan Kult,
head of engineering with Riddell, in Chicago, which makes protective equipment
including helmets for the sport. 鈥淲e are always looking for ways to get a
realistic picture of an athlete wearing one of our helmets.鈥
Kult currently tests Riddell helmets using a dummy head similar to those in
car crash tests, but he is particularly interested in DERAMan鈥檚 ability to pick
up the shear waves caused by a glancing blow in a collision. 鈥淲e just don鈥檛
understand rotational injuries,鈥 he says. 鈥淲e would love to know more about
their effects on the brain.鈥
Shaken and stirred
Meanwhile, Alan Hepper, a specialist in ejection seats at DERA, wants to use
the dummy head to explore what happens to the brain when a pilot ejects at high
speed from the cockpit of a fighter plane. During ejection, pilots first feel a
force of 16g which may cause their head to hit their knees. When they
leave the cockpit, the rush of air outside the plane flings their head back
against the head box. So despite the helmets and padding round the head box, the
brain takes a good shaking. Hepper wants DERAMan to help him solve the problem
of building an ejection system which not only prevents brain damage but also
ensures that the pilot stays conscious throughout the ordeal and during the
landing. 鈥淏y using a smarter dummy we could test our predictions of the
probability of head injury and make some more intelligent modifications to the
design of the seats,鈥 he says.
Hepper is also keen for DERAMan to be given a spine because this is a common
site of injury in pilots who are forced to eject. 鈥淲e want to tune the speed
that the ejector seat blasts off,鈥 he says. 鈥淭he last thing we want to do is to
send a shock wave travelling up the pilot鈥檚 spine with a wavelength that
reflects off the top of their head and sets up a standing wave in their
蝉辫颈苍别.鈥
At the moment, there is no way of testing for this. But a spine built using
plastic vertebrae and discs with pressure gauges in each would give Hepper
unprecedented insights into the stresses suffered by a human body being shot out
of a plane. Bell even says that it would be possible to build a whole body using
the technology behind DERAMan. But for most applications, she believes it is
more likely that sections of smart dummy will be built into existing 鈥渄umb鈥
dummies.
Bell has only just launched the DERAMan project, but she is confident that
her sensitive head will find itself being bolted onto a broad range of shoulders
in the future. She hopes that putting dummies in touch with some of their
feelings will pay off in the form of safer cars, safer sports, safer ejector
seats and ultimately a design tool to highlight and prevent head injuries.
DERAMan could soon be making鈥攁s well as taking鈥攁 big impact.

- Further reading: 鈥淐linical and neuropsychological aspects of closed head
injury鈥 by John Richardson, Taylor and Francis 1990.