Title: Helmet Protection against Traumatic Brain Injury: A Physics Perspective
1Helmet Protection against Traumatic Brain Injury
A Physics Perspective
Eric Blackman University of Rochester
Acknowledgments -DSSG program,Institute for
Defense Analyses -Melina Hale (Chicago)
-Sarah Lisanby (Columbia) -Willy Moss (LLNL)
-Michael King (LLNL)
22
33
44
55
6Closed Traumatic Brain Injury (TBI)
- physical injury to the brain without skull
fracture - concussions (non-local midbrain brainstem
frontal lobe) - diffuse axonal injury (shear damage of axons,
white matter grey matter linkage) - contusions (general bruising)
- subdural hematoma (bridging vein damage)
6
7Sources of TBI without skull fracture
- head impacts (ITBI)
- skull comes to a stop, brain crashes into skull
- blast overpressure (OTBI)
- brain deforms from pressure and shear waves even
without bulk acceleration (skull flexure--more
later) - possible alternate route via thorax to brain
- blasts propel victim so ITBI OTBI must be
common (B07)
8Cost of TBI (USA)
- Human costs
- Civilian 2x106 cases/yr 50 auto 25 sports
(McArthur 04) - 20 deaths per 100,000 20 billion/yr treatment
- Military
- before 2006 estimated 3 of soldiers have TBI
(60 of hospital injured soldiers) - 0.6 of all solidiers serious TBI
- New screenings 2006-2009 18 of all troops
have TBI (Fts. Carson, Hood, Bragg, Cp.
Pendelton..),1.5 of all troops unfit to return. - 2.7 million (Blimes 07) per 25 yr post-TBI
lifetime gt2 billion/year for treatment - Workforce / mission / security costs
9Why is TBI a Research Frontier? Modern
equipment has reduced fatalities, leaving
survivors with prevalent secondary injuries.
Many aspects of TBI science are therefore nascent
- medical screening and correlation with trauma
- macho culture TBI not always understood as
physical - PTSD vs TBI
- treatment
- physiology and biology of injury
- mechanisms of injury (impact vs. blast)
- physics of protection
9
10Head Impacts
- Gravity or explosion converts gravitational
potential energy or chemical energy into bulk
kinetic energy - Rapid deceleration upon impact implies large
force - During impact, kinetic energy is converted into
deformation energy - Brain damage from energy dissipated in brain
rather than helmet or skull - tissue stress (force per unit area) threshold for
injury - duration of force threshold for injury
- Functional forms not well understood
11TBI from Impacts
- As head impacts, brain keeps moving it is
coupled to skull by cerebral spinal fluid (CSF) - Brain crashes into skull displacing fluid
stresses brain tissue both by compression and
shear - Protecting skull from fracture is insufficient to
protect brain from crashing into skull - Need to
- reduce head acceleration (reduces maximum force
incurred by brain-skull crash) - reduce energy absorbed by brain (reduces energy
available to sustain a distorted brain for
extended period)
12Role of Helmets for ITBI
- Hard shell protects the skull at point of impact
and absorbs, redistributes, and releases energy
rapidly - Hard shell alone is no good
- Need cushioning to reduce head impact
acceleration and thus force on brain - Cushioning standard must be more stringent to
protect against brain injury than skull fracture - subtleties in helmet/skull/brain force coupling
13Origin of TBI/Blunt Impact Standards
Ono et al. 1980 (human cadaver and scaled monkey
data)
1414
15Impact Acceleration Profile
Peak force for short time
Lower force over longer time
16Widely used Injury Measures
1. Peak g
2.
3.
- SI and HIC empirically accommodate acceleration
and duration from cadaver and animal injury data
- Can create human injury probability graph
- e.g. Head HIC gt 1000 (sec), 16 risk of life
threatening TBI (Prasad Mertz 1985) scaled
monkey data auto industry.
17Classification of TBI Severity (Hayes et al. 07)
17
18HIC15 AIS4 Injury Risk (Prasad Mertz 85, data
compilation)
19How are HIC and SI used?
- NHTSA uses HIC 1000. (supposedly 1 chance of
fatality 30MPH collision for restrained driver) - NOCSAE uses SI1200 ( JHTC) but for NFL does
not fully protect against TBI should be 140
based on data - NO SI/HIC standard for military Helmets peak g
standard only and its NO GOOD. - Slobodnik (1980) need lt150G at 1.5 meters drop
- special forces helmets standard is 150G at 1.5
feet(!) - Free falls of 3 feet for a 5kg head form
including PAGST or ACH helmets give 300G
(McEntire et al.05)
20(No Transcript)
21(No Transcript)
22(No Transcript)
23Current Military Helmets Fail
Wayne State Head Tolerance (WSHTC) Gujdardan et
al. 1966
Combat Helmets for 4.5 ft drop
McEntire et al. 05
- Japan Head Tolerance (JHTC, 25 p of concussion
HIC1100) Ono et al. (1980,scaled monkey data)
24(Pellman et al. 03)
Viano et al 2007
25Innovative use of Accelerometers (Duma et al. 05
(03 season) college football)
26(No Transcript)
27 Military helmets 4.5 foot drop
30 risk curve like JHTC (uses scaled monkey
data)
- Actual gt70
- NFL risk curve
Blast simulations (w/injury) (Moss, King,
Blackman 09)
augmented from Pellman et al (03,06)
28ITBI protection standards AND measures are
flawed
- measures HIC, SI based only on (limited)
experimental data body mass and impact angles
not included, have little theoretical foundation,
not even the best indicators... - even if measures were correct standards in
military and NFL are inadequate - ITBI measures are useless for BTBI (later)
28
29JHTC
Woodpeckers dont get TBI HIC relies on fixed
brain mass and surface area
30(No Transcript)
31Role of Body Mass and Impact Angle on Injury
Thresholds (Blackman 2009)
31
32Physical Quantities that TBI measures should
correlate
- linear force (mass X linear acceleration)
- total energy and energy input rate
- torque (moment of inertia X rotational
acceleration) - Internal
- brain tissue stress or pressure maximum
- brain tissue rate of elastic energy change
(localized)
33Numerically Simulating Impact Head ModelsTBI
Thresholds based on Internal Forces
- Zhang et al. 04 reproduced NFL collisions with
Wayne State Head Model - WSHM gray matter (cell) white (fibrous) shear
moduli 20 larger for white white is 2-D
isotropic, grey is 3-D isotropic brain stem
shear mod 40 higher than cerebrum etc.. - Data on these properties differ, but code can
incorporate whatever the data require
34TBI internal measures from simulations
- Zhang et al 04 reproduce videoed impacts with
head form collisions, then use experimental data
as input for numerical simulations to calculate
internal stresses - Maximum stress at core (diencephalon, upper brain
stem) - rate of maximum strain ( rate of elastic energy
change) and peak stress are best correlators with
injury
35Coup Contrecoup pressures
36(Pellman et al. 03)
Viano et al 2007
37NFL MTBI measures (King 03 Zhang et al. 04
Newman 05 )
38P versus peak Linear Acceleration
P versus HIC
Football HIC no better than Peak Acceleration or
Impact Power as measure
P versus Impact Power
Newman 05 King 03
39NFL Data/Analysis With Drop Tests Suggest Better
Impact TBI Measures can be Identified
- Newman 2000 e.g. best external correlators
- impact power
- peak linear acceleration
- King et al 05 Liang et al. 04 e.g. best
internal correlators are - Rate of maximum elastic energy change
- Maximum stress
- goal of TBI protection design should be to
robustly understand how external forces correlate
with internal stresses and injury
40- Left helmet has reduced concussions by 31
relative to right helmet in high school football
(Collins et al. 2006)
41(No Transcript)
42Military data are needed
- Unreliability of head indices and different
nature of impacts in football, auto, and combat - Mass matters, neck strength matters
- Overpressure injury involves different
mechanisms than impact injury - Blast produces pressure impact injury
43Blast Injury Protection
- frontier of helmet design
- Sources of Injury
- Primary (overpressure)
- Secondary (shrapnel)
- Tertiary (impact)
43
44Basic Blast Physics
Static over-pressure at fixed position
45Basic Blast Physics
46Blast Wave Injury
(Moss King, personal comm.)
47Outdated Bowen Curves No TBI threshold
200 atm
1 atm
100 atm
1 atm
48 Initial Simulations of Blast vs. Impact
Moss, King, Blackman (2009)
49The Head in the LLNL Sims.
50(2.3 kg C4)
50
51Model for Impact
- HIC 1090
- peak g 194 g
- impact duration 2.1 ms
51
52Snapshot of Impact vs Blast Pressures
52
53(No Transcript)
54- Reducing wave speed in brain by reducing bulk
modulus produces deeper penetration of pressure
extremes as stress gradients are slower to relax
- Brain pressure peaks then followed by after
shocks long after blast front passes
55(No Transcript)
56(No Transcript)
57(No Transcript)
58Role of Current Helmets for Blast
- without pads, underwash amplfies pressure
under helmet - but with pads, head is more strongly coupled to
skull - need to optimize for blast impact
59Moss et al. 09 .
60Initial Lessons From Blast-Head Interaction
Simulations
- Skull flexure NOT acceleration is primary
mechanism of OTBI through skull - need helmet that prevents skull flexure a rigid
shell foam that damps the stress waves - underwash in current helmets exacerbates injury
- current helmet cushioning protects against
underwash but does not damp skull flexure - need to optimize OTBI and ITBI protection
- skull may not be only pathway
61Future Studies
- BLAST overpressure (OTBI)
- Add more realistic head model
- maybe skull is not path of maximum coupling to
brain - blast inside enclosure (vehicle)
- IMPACT (ITBI)
- Include body attached to head for the impact and
vary impact with angle to extract effective mass - revise HIC to include impact angle body mass
- For BOTH
- Correlate specific external forces with specific
internal stresses - Run impact simulations for pre-injured brain from
overpressure - Correlate specific brain damage with medical
symptoms - Optimize helmets
62END
62
63Properties of Simulated Skull/Brain/Helmet Foam
- Skull Youngs modulus9Gpa, density 1.7g/cc
- brain density 1.04 g/cc, instantaneous shear
modulus 37.5kPa quasi static shear
modulus6.95kPa, decay factor 700 /s bulk
modulus2.19Gpa (water) - helmet foam bulk mod1.3Mpa, quasi static shear
mod20.1 kPa instantaneous shear mod2 MPa,
decay factor 100/s, density 0.136g.
63
6464
65Present Helmets Ineffective Against TBI but Can
be Improved
- Blast causes impact injury and overpressure
injury - Need better TBI measures and standards
- Short Term cushioning inserts for that reduce
head form acceleration by factor gt 3 from 5 ft.
drop - Longer Term Present TBI helmet measures are
imprecise - Mass is not included in current impact TBI
measures - need combat specific data (accelerometers,
pressure gauges) to correlate with clinical TBI
for blast and impact - numerical simulations of blast and impact with
head model - correlate external forces with internal brain
stresses and symptoms - NFL/NCAA Football lead way for use of technology
to improve helmet design and TBI monitoring from
impacts - Overpressure safety threshold needed NOT based
on accel. cushioning helps but worst injury may
not be via skull - Progress from blast impact simulations (w/Moss
King, LLNL)
66Present Combat Helmets Ineffective Against TBI
but Can be Improved
- Blast causes both impact injury and overpressure
injury - Need better TBI measures and tougher standards
- Short Term cushioning inserts for that reduce
headform acceleration by factor gt 3 from 5 ft.
drop - Longer Term Present TBI helmet measures are
imprecise - Mass is not included in current impact TBI
measures - need combat specific data (accelerometers,
pressure gauges) to correlate with clinical TBI
for blast and impact - Need numerical simulations of blast and impact
with head model correlate external forces with
internal brain stresses and symptoms - NFL/NCAA Football lead way for use of technology
to improve helmet design and TBI monitoring from
impacts - Overpressure safety threshold needed NOT based
on bulk acceleration cushioning helps BUT is
worsts injury via skull? - Progress and promise from blast impact
simulations (w/ Moss King LLNL)
67Conclusions/Recommendations
- Many costs of TBI gt1.1 billion/yr for mil.
medical care - Can make substantial improvements
- Blast is a 2 stage injurer overpressure and
impact - Both need to be addressed
- Improving impact TBI measures and standards
- Quick fix use 1/3 blunt acceleration standard
for MICH helmets and get new inserts for ALL
combat helmets. - Pad the inside of vehicles
- Longer term Field data from accelerometers
coupled with numerical simulations to correlate
external and interal forces - Directed funding specifically to improve impact
injury TBI measures and standards directed
toward identifying the optimal cushioning inserts
68Conc./Recommendations (cont)
- Effective mass often greater than head mass
derive from theory and field data lowers current
thresholds for injury - HIP better than widely used HIC
- Test bivariate correlation HIP and Amax with
injury and correlate with internal stresses from
simulations - Improving overpressure TBI measures standards
- Need field data pressure gauges in helmet and
on body - overpressure TBI thresholds from correlating
external forcing with internal stress in models
and comparison to field data - overpressure measures are NOT based on bulk
acceleration so HIC etc are useless. - Same cushioning from impact will help but need to
how blast overpressure is transmitted to brain
simple models can help
69Conc./Recommendations (cont)
- aerodynamic issues, passive vs. active blast
mitigation - neck and tethering
- blast experiments for TBI
- Great promise from numerical simulations of blast
impact Ultimately correlate external forces
with internal brain tissue stresses and medical
symptoms - tackling the most pertinent questions can reduce
complexity
70(No Transcript)
71Example Blast Simulation
MKB 09
7272
73Pellman et al. 03 (Viano, Pellman,
Witherall..03-07)
74HIC and SI are flawed
- No dependence on mass
- Not the best indicator of impact injury even when
mass is constant - Not quantitatively derived from first
principles - based on empirical formulae that fit un-refereed
scaled monkey data (JHTC 80) and cadaver skulls
(WSHTC, 61) - Doesnt distinguish specific injury type well
- What about neck strength? (note HIC ?V4/ d1.5)
Chou Nyquist(74) - HIC and SI are useless for overpressure blast
injury
75Use of Accelerometers in NCAA
- 6 accelerometers measure all head motion store
data send data to sideline laptop - IMPACT or other software to clinically diagnose
MTBI - Correlate data with incidence of TBI
- Input into head model to identify internal brain
tissue stress - Ironically VT study was of limited use because
the helmets are too GOOD (only 1 concussion from
3000 impacts monitored) - Need to adopt for military use in combo with
pressure measurements for blast correlation
76(No Transcript)
77Helmets and ITBI standards
- Current combat helmets no TBI protection
standard - NOCSAE (sports) yes, NHTSA (motor vehicle)
(fatality TBI only) - Special forces MICH helmets have a required blunt
impact standard but it insufficient - ACH and PASGT (infantry helmets) with padding
kit meet MICH standard and thus also
insufficient. - What is reasonable standard? (fundamental
question) - Bare Minimum Slobodnik (1980) 150G peak
acceleration at 1.5 meters drop needed from
reconstructing crashed helicopter injuries - MICH 150G at 1.5 feet (FAILS)