USF Libraries
USF Digital Collections

Evaluation of advanced materials to protect against fall-related head injuries

MISSING IMAGE

Material Information

Title:
Evaluation of advanced materials to protect against fall-related head injuries
Physical Description:
Book
Language:
English
Creator:
Kerrigan, Michael V
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Falls
Biomechanics
Accelerometer
Head Injury Criteria (HIC)
Traumatic Brain Injury (TBI)
Dissertations, Academic -- Chemical Engineering -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Falls among the elderly population continue to be a growing concern in the healthcare industry and are marked by staggeringly high social and economic costs. The incidence of falls is known to increase with age, and currently the elderly population is growing at an astounding rate as baby-boomers are now entering this age group. Also, recovery following fall-related injuries decreases with increased age. These confounding factors currently make falls a very important area of research. Of the injuries typically seen in falls among the elderly, head injuries are one of the most debilitating. Death due to head trauma among the elderly is gaining national attention; head trauma is now considered the number one cause of death among elders who fall1. Among other technologies, medical helmets are often employed to protect against such injuries, but patient compliance with these helmets remains an issue. Current helmets use foams and cotton as padding, contributing to clumsy designs. Dilatent and honeycomb materials may be the future of this industry as their low weight and high efficacy per thickness make them ideal materials for thinner, lighter, less cumbersome head protection devices. This study outlines various modes of head injury and then highlights several head protection measures. The newer materials are tested using various methods to determine the most promising candidates for prototype designs. Next, three prototypes are assembled from the newer materials and compared directly based on the protection measures established. Finally, the top-performing prototype is compared against two existing medical helmets in a similar fashion. The results show that the best prototype significantly outperforms one of the existing medical helmets, and shows slight improvement over the other. These results establish the promise of these newer materials in the application of head protection devices.
Thesis:
Thesis (M.S.B.E.)--University of South Florida, 2009.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Michael V. Kerrigan.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 133 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002069237
oclc - 608021951
usfldc doi - E14-SFE0003103
usfldc handle - e14.3103
System ID:
SFS0027419:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Evaluation of Advanced Materials to Protect Against Fall Related Head Injuries by Michael V. Kerrigan A thesis submitted in partial fulfillment o f the requirements for the degree of Master of Science in Biomedical Engineering Department of Ch emical Engineering College of Engineering University of South Florida Co Major Professor: John D. Lloyd, Ph D Co Major Professor: William E. Lee III Ph D Member: Tat jana Bulat, M D Date of Approval: June 18, 2009 Keywords: falls, biomechani cs, accelerometer, Head Injury Criteria (HIC), Traumatic Brain Injury (TBI) Copyright 2009, Michael V. Kerrigan

PAGE 2

DEDICATION This work is dedicated to my loving family for their continuous encouragement throughout my academic career. I s urely would not have made it this far without their love and support. Thank you Mom, Dad, Di, Rick, K.C., Carly, Cara, and Lola!

PAGE 3

! TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ............. iv LIST OF FIGURES ................................ ................................ ................................ ........... vi LIST OF EQUATIONS ................................ ................................ ................................ ..... xi ABSTRACT ................................ ................................ ................................ ...................... xii INTRODUCTION ................................ ................................ ................................ ............... 1 LITE RATURE REVIEW ................................ ................................ ................................ .... 3 Introduction ................................ ................................ ................................ ............. 3 Fall Frequency in the Elderly Population ................................ ............................... 3 Rate of Falling in Persons 65+ Years of Age ................................ ............. 3 Growth of the Elderly Population ................................ ............................... 4 Head Injuries in the Elderly ................................ ................................ .................... 4 Basic Anatomy of the Intracranial Cavity ................................ .................. 4 Common Head Injuries Experienced During Falls ................................ ..... 5 Traumatic Brain Injury ................................ ................................ ... 5 Skull Fracture ................................ ................................ .................. 7 Risk Factors ................................ ................................ ................................ 7 Blood Thinners ................................ ................................ ................ 7 Age Related Factors ................................ ................................ ......... 7 Muscle Strength Degradation ................................ .............. 8 Brain Atrophy ................................ ................................ ...... 9 Prognosis of Fall Related Head Injuries in the Elderly ....... 9 Social and Economic Costs ................................ ................................ ..................... 9 HELMET TESTING STANDARDS ................................ ................................ ................. 11 Anthropomorphic Test Dummy (ATD) Headform ................................ ................ 11 Bio fidelity ................................ ................................ ................................ 11 Head Protection Measures ................................ ................................ .................... 12 Head Injury Criterion ................................ ................................ ................ 12 Peak Force ................................ ................................ ................................ 15 Peak Resultant Translational Acceleration ................................ ............... 16 Peak Rotational Acceleration ................................ ................................ .... 16 Gadd Severity I ndex ................................ ................................ .................. 16 Headform Drop Testing Methods ................................ ................................ ......... 17 Guided Drop Tests ................................ ................................ .................... 17 Twin Wire Test Apparatus ................................ ............................ 17 Monorail Test Apparatus ................................ .............................. 18 Limitations ................................ ................................ .................... 18

PAGE 4

! "" Freefall Drop Tests ................................ ................................ ................... 18 Platform Test Apparatus ................................ ............................... 18 True Freefall Test Apparatus ................................ ........................ 18 Limitations ................................ ................................ .................... 19 Comparison of Guid ed and Freefall Tests ................................ ................ 19 Various Industries Studied ................................ ................................ .................... 19 Medical Helmets ................................ ................................ ....................... 20 Motorcycle Helmet Standards ................................ ................................ .... 20 Snell Standard ................................ ................................ ................ 20 DOT Standard ................................ ................................ ................ 21 BSI Standard ................................ ................................ .................. 21 ECE Standard ................................ ................................ ................. 21 Soccer Helmet Standards ................................ ................................ ........... 22 Playground Surfacing Materials Standards ................................ ............... 22 Rugby Helm et Standards ................................ ................................ .......... 23 Bicycle Helmet Standards ................................ ................................ ......... 23 Conclusion ................................ ................................ ................................ ............. 23 METHODS ................................ ................................ ................................ ....................... 24 Int roduction ................................ ................................ ................................ ............ 24 Location ................................ ................................ ................................ ................. 24 Preliminary Head Drop Testing ................................ ................................ ........... 24 Apparatus Design and Construction ................................ .......................... 24 Protocol ................................ ................................ ................................ ...... 27 Data Collection, Calibration, and Verification ................................ ......... 28 Phase I: Impact Testi ng ................................ ................................ .......................... 32 Apparatus Design and Construction ................................ .......................... 32 Protocol ................................ ................................ ................................ ...... 34 Data Collection ................................ ................................ .......................... 34 Data Processing ................................ ................................ .......................... 34 Phase II: Head Drop Testing ................................ ................................ .................. 35 Apparatus Design and Construction ................................ .......................... 35 Protocol A ................................ ................................ ................................ .. 35 Protocol B ................................ ................................ ................................ .. 35 Data Collection ................................ ................................ .......................... 36 Data Processing ................................ ................................ .......................... 36 RESULTS ................................ ................................ ................................ .......................... 37 Phase I Results ................................ ................................ ................................ ....... 37 Phase II Results ................................ ................................ ................................ ...... 40 DISCUSSION ................................ ................................ ................................ ................... 67 Interpretation ................................ ................................ ................................ .......... 67 Limitations ................................ ................................ ................................ ............ 68 CONCLUSIONS AND FUTURE RESEARCH ................................ ............................... 69 LITERATURE CITED ................................ ................................ ................................ ..... 71

PAGE 5

! """ APPENDIX 1: MATLAB CODE ................................ ................................ ...................... 74 APPENDIX 2: SUPPLEMENTARY RESULTS ................................ .............................. 92 APPENDIX 3: PHOTOGRAPHS OF MATERIALS & MEDICAL HELMETS ........... 110

PAGE 6

! "# LI ST OF TABLES Table 1: Risk levels with associated head injury measures ................................ ............... 22 Table 2: Determined conversion factors for indicated force sensors ................................ 30 Table 3: Results of head drop tes ting of individual materials and materials in combination ................................ ................................ ................................ ......... 41 Table 4: Results of prototype comparison (back orientation) ................................ ............ 51 Table 5: Results of prototype comparison (front orientation) ................................ ........... 52 Table 6: Results of prototype comparison (side orientation) ................................ ............. 53 Tab le 7: Results of helmet comparison (back orientation) ................................ ................ 58 Table 8: Results of helmet comparison (front orientation) ................................ ................ 59 Table 9: Results of helmet comparison (side orientation) ................................ ................. 60 Table 10: Results of comparison of "Prototype 1" and the unprotected case (bac k orientation) ................................ ................................ ................................ ........ 64 Table 11: Results of comparison of prototype and the unprotected case (front orientation) ................................ ................................ ................................ ....... 65 Table 12: Results of comparison of prototype and the unprotected case (side orientation) ................................ ................................ ................................ ....... 66 Table 13: Impact testing results of individu al materials ................................ .................. 101 Table 14: Impact testing results of material combinations ................................ .............. 102 Table 15: Head drop testing res ults of the unprotected case ................................ ........... 103 Table 16: Head drop testing results for "Prototype 1" ................................ ..................... 105 Table 17: Head drop testing results for "Prototype 2" ................................ ..................... 1 06 Table 18: Prototype 3 head drop testing results ................................ ............................... 107

PAGE 7

! # Table 19: Head drop testing results for Plum ProtectaCap+Plus medical helmet ........... 108 Table 20: Head drop testing results for HipSaver medical helmet ................................ .. 109

PAGE 8

! #" LIST OF FIGURES Figure 1: Wayne State Tolerance Curve (WSTC) ................................ ............................. 13 Figure 2: Head drop testing tower ................................ ................................ ..................... 27 Figure 3: Verification of Chatillon gauge accuracy ................................ ........................... 28 Figure 4: Verifica tion technique for Chatillon gauge ................................ ........................ 29 Figure 5: Example of conversion factor determination (force sensor A) .......................... 30 Figure 6: Compression test with Chatillon gauge on force sensor ................................ .... 30 Figure 7: Schematic of data collection ................................ ................................ ............... 31 Figure 8: Acceleration vs. time traces for dilatent materials tested on the impacting tower. ................................ ................................ ................................ ................. 38 Figure 9: Acceleration vs. time traces for honeycomb materials tested on the impacting tower. ................................ ................................ ................................ ................. 39 Figure 10: Head drop testing results for the unprotected case (peak acceleration). .......... 44 Figure 11: Head drop testing results for the unprotected case (HIC15). ........................... 45 Figure 12: Head drop testing results for the unprotected case (peak for ce). ..................... 46 Figure 13: Prototype comparison (peak acceleration). ................................ ...................... 48 Figure 14: Prototype comparison (peak force). ................................ ................................ 49 Figure 15: Prototype comparison (HIC15). ................................ ................................ ....... 50 Figure 16: Comparison of prototype, Plum, and HipSaver helmets (peak acceleration in the front orientation). ................................ ................................ ................... 55 Figure 17: Comparison of prototype, Plum, and HipSaver helmets (peak force in the front orientation). 56 !

PAGE 9

! #"" Figure 18: Comparison of prototype, Plum, and HipSaver helmets (HIC15 in the front orientation). ................................ ................................ ................................ ...... 57 Figure 19: Comparis on of prototype 1 and unprotected case (HIC15 in the front orientation). ................................ ................................ ................................ ...... 62 Figure 20: Comparison of prototype 1 and the unprotected case (peak acceleration in the front orientation). ................................ ................................ ................... 63 Figure 21: Comparison of 3 prototype designs (peak accele ration in the b ack orientation) ................................ ................................ .............................. 93 Figure 22: Comparison of 3 prototype designs (peak acceleration in the side orientation) ................................ ................................ ............................... 93 Figure 23: Compariso n of 3 prototype designs (peak force in the back orientation) ........ 94 Figure 24: Comparison of 3 prototype designs (peak force in the side orientation) .......... 94 Figure 25: Comparison of 3 prototype designs ( HIC15 in the back orientation) .............. 95 Figure 26: Comparison of 3 prototype designs (HIC15 in the side orientation) ............... 95 Figure 27: Comparison of "Prototype 1" and 2 medical helmets (peak acceleration in the back orientation) ................................ ................................ ........................ 96 Figure 28: Comparison of "Prototype 1" and 2 med ical hel mets (peak acceleration in the side orientation) ................................ ................................ .......................... 96 Figure 29: Comparison of "Prototype 1" and 2 medical helmets (peak force in the back orientation) ................................ ................................ ................................ ....... 97 Figure 30: Comparison of "Prototype 1" and 2 medical helmets (peak force in the side orientation) ................................ ................................ ................................ ....... 97 Figure 31: Comparison of "Prototype 1" and 2 medical helmets (HIC15 in the back orientation) ................................ ................................ ................................ ....... 98 Figure 32: Comparison of "Prototype 1" and 2 medical helmets (HIC15 in the side orientation) ................................ ................................ ................................ ....... 98 Figure 33: Comparison of "Prototype 1" to the u nprotected case (HIC15 in the back orientation) ................................ ................................ ................................ ....... 99 Figure 34: Comparison of "Prototype 1" to the unprotected case (HIC15 in the side orientation) ................................ ................................ ................................ ....... 99

PAGE 10

! #""" Figure 35: Comparison of "Prototype 1" to the unprotected case (peak acceleration in the back orientation) ................................ ................................ ...................... 100 Figure 36: Comparison of "Prototype 1" to the unprotected case (peak acceleration in the side orientation) ................................ ................................ ........................ 100 Figure 37: "DC W" dilatent material (front view) ................................ ........................... 111 Figure 38: "DC W" dilatent material (back view) ................................ ........................... 111 Figure 39: "A" dilatent material (front view) ................................ ................................ .. 112 Figure 40: "A" dilatent m aterial (back view) ................................ ................................ ... 112 Figure 41: "S2" dilatent material, also referred to as "Dd" (front view) ......................... 113 Figure 42: "S2" dilatent material, also referred to as "Dd" (back view) .......................... 113 Figure 43: "B" dilatent material (front view) ................................ ................................ ... 114 Figure 44: "B" dilatent material (back view) ................................ ................................ ... 114 Figure 45: "C" dilatent material (front view) ................................ ................................ ... 115 Figure 46: "C" dilatent material (back view) ................................ ................................ ... 115 Figure 47: "DC R" dilatent material (front view) ................................ ............................ 116 Figure 48: "DC R" dilatent material (back view) ................................ ............................ 116 Figure 49: "S7" di latent material (front view) ................................ ................................ 117 Figure 50: "S7" dilatent material (back view) ................................ ................................ 117 Figure 51: "DC Y" dilatent material (front view) ................................ ............................ 118 Figure 52: "DC Y" dilatent mat erial (back view) ................................ ............................ 118 Figure 53: "D1" dilatent material (front view) ................................ ................................ 119 Figure 54: "D1" dilatent material (back view) ................................ ................................ 119 Figure 55: "F25" honeycomb material, also r eferred to as "Hz" (front view) ................. 120 Figure 56: "F25" honeycomb material, also referred to as "Hz" (back view) ................. 120

PAGE 11

! "$ Figure 57: "K" honeycomb material (front view) ................................ ............................ 121 F igure 58: "K" honeycomb material (back view) ................................ ............................ 121 Figure 59: "F33" honeycomb material (front view) ................................ ........................ 122 Figure 60: "F33" honeycomb material (back view) ................................ ......................... 122 Figure 61: D" honeycomb material, also referred to as "Hy" (front view) .................... 123 Figure 62: "D" honeycomb material, also referred to as "Hy" (back view) .................... 123 Figure 63: "L" honeycomb material (front view) ................................ ............................ 124 Figure 64: "L" honeycomb material (back view) ................................ ............................ 124 Figure 65: "J" honeycomb material (front view) ................................ ............................. 125 Figure 66: "J" honeycomb material (back view) ................................ ............................. 125 Figure 67: "H" honeycomb material (front view) ................................ ............................ 126 Figure 68: "H" honeycomb material (back view) ................................ ............................ 126 Figure 69: "E" honeycomb material (front view) ................................ ............................ 127 Fig ure 70: "E" honeycomb material (back view) ................................ ............................ 127 Figure 71: "I" honeycomb material (front view) ................................ ............................. 128 Figure 72: "I" honeycomb material (back view) ................................ .............................. 128 Figure 73: "G" hon eycomb material (front view) ................................ ............................ 129 Figure 74: "G" honeycomb material (back view) ................................ ............................ 129 Figure 75: "F25E" honeycomb material (front view) ................................ ...................... 130 Figure 76: "F25E" honeycomb m aterial (back view) ................................ ...................... 130 Figure 77: "D2" dilatent material (front view) ................................ ................................ 131 Figure 78: "D2" dilatent material (back view) ................................ ................................ 131 !

PAGE 12

! $ Figure 79: HipSaver medical helmet ................................ ................................ ............... 132 Figure 80: Plum ProtectaCap+Plus medical helmet ................................ ........................ 133

PAGE 13

! $" LIST OF EQUATIONS Equation 1: Root mean square of acceleration data ................................ ........................... 13 Equation 2: Head Injury Criterion (HIC) formula ................................ ............................. 14 Equation 3: Gadd Severity Index (GSI) formula ................................ ............................... 17 Equation 4: Pressure formula ................................ ................................ ............................. 32

PAGE 14

! $"" Evaluat ion of Advanced Materials to Protect Against Fall Related Head Injuries Michael V. Kerrigan ABSTRACT Falls among the elderly population continue to be a growing concern in the healthcare industry and are marked by staggeringly high social and econom ic c osts. The incidence of falls is known to increase with age, and currently the elderly population is growing at an astounding rate as baby boomers are now entering this age group. Also, recovery following fall related injuries decreases with increased age. These confounding fact ors currently make fall s a very important area of research Of the injuries typically seen in falls among the elderly, head injuries are one of the most debilitating. Death due to head trauma among the elderly is gaining nati onal attention; head trauma is now considered the number one cause of death among elders who fall 1 Among other technologies, medical helmets are often employed to protect against such injuries, but patient compliance with the se helmets remains an issue. Current helmets use foams and cotton as padding, contributing to clumsy designs. Dilatent and honeycomb materials may be the future of this industry as their low weight and high efficacy per thickness make them ideal materials for thinner, lighter, less cumbersome head protection device s This study outlines various modes of head injury and then highlights several head protection measures The newer materials are tested using various methods to determine the most promising candidates for prototype des igns. Next, three prototypes are assembled from the newer materials and compared directly based on the protection measures established. Finally, the top performing prototype is compared against two existing medical helmets in a similar fashion. The resu lts show that the best prototype significantly outperforms one of the existing medical helmets, and shows slight

PAGE 15

! $""" improvement over the other. These results establish the promise of these newer materials in the application of head protection devices.

PAGE 16

! % INTRODUCTION Falls among the elderly represent a major concern in today's healthcare industry and require significant attention in order to maximize the outcomes for those who do suffer from falls. Much effort has been dedicated to reducing the instanc e of falls and improving the resulting consequences, including training programs for healthcare staff, wandering alarms to alert staff that a patient is walking about, bedside floor mats to reduce injuries due to falls, and medical helmets to reduce severi ty of injury to the head in the event of a fall. M edical helmets provide a unique solution to this issue in that they can directly reduce those physical measures believed to cause head injuries such as accelerations and forces acting on the head during the head's impact with another object (i.e. the floor). Reducing the magnitude of these measures is a practical way of minimizing injury in the event of a fall. Several medical helmets are available for prescription to those thought to be at high risk of a fall or those who have just suffered various types of head injuries. The current designs are often clumsy in appearance, introducing the issue of patient compliance with their prescription. Future work in this field should be aimed at appearance, size weight, and any other factors believed to increase patient compliance. Dilatent and honeycomb materials may be the future of this industry as their low weight and high efficacy per thickness make them ideal materials for thinner, lighter, less cumbers ome head protection device s Dilatent material is designed to stiffen upon the sudden application of an external force. This characteristic makes it potentially useful to the medical helmet application as it can become rigid in the event of a collision o f the head and shunt the impacting force over a greater area of the head. Spreading the external force over a greater area reduces stress on the skull, lowering the risk of fracture. Honeycomb material acts much like foam by deforming to reduce the amoun t of force transferred to the head by the impacted object. For this application it is potentially better than foam because it is very lightweight, leading to a lightweight helmet and potentially

PAGE 17

! & increasing patient compliance. These materials have previou sly not been adapted to this field, and the current study aimed to evaluate their potential use in medical helmets. Assessing the effectiveness of these materials was be accomplished by borrowing concepts used in existing helmet testing methods to create an original testing method. A group of honeycomb and dilatent material samples were tested individually and in combination using multiple methods to assess their impact attenuation characteristics. The results were used to determine those materials or material combinations with the greatest potential for application to medical helmets. Based on these findings, three prototypes were constructed and compared directly to two existing medical helmets using an original drop testing method.

PAGE 18

! LITERATURE REVIEW Introduction The current study focuses on assessing and applying head protection to an elderly population, which requires special consideration due to factors that are outlined in the following review of the literature. The firs t topic covered is the increased frequency of falls among the elderly. Next, the types and modes of head injury commonly associated with falls in the elderly are covered. With a foundation established on the issue of falls among the elderly, the literatu re review turns towards current helmet testing standards, covering a broad range of industries and populations. This provides the basis for the design of our testing procedures and assessment of the efficacy of various head protection devices. Fall Frequ ency in the Elderly Population Rate of Falling in Persons 65+ Years of Age Approximately one third of all adults over age 65 and one half of adults over age 80 are reported to fall each year [2, 3] "Approximately 8% of persons aged 65 and older visit t he emergency department each year because of fall related injury, with about 25% of these visits res ulting in inpatient admission" [4] One study found that between "1996 and 1999, 71% of fatal fall related traumatic brain injuries ( TBI ) occurred in adults aged 65 and older" [ 4 ]. Once hospitalized, older adults fall about 1.5 times per bed per year. Also, it has been shown that the instance of "a single fall is a major risk factor for a subsequent fall, increasing the risk of repetitive TBI" [ 4 ].

PAGE 19

! ( Gr owth of the Elderly Population The number of persons aged over 65 in the U.S. is expected to double from the 35 million as of 2006 to 70 million by the year 2030 [ 4 ]. This increase in the elderly population of the U.S. is strongly linked to the baby boom following World War II, which started about 65 years ago. Another study estimated that the worldwide population aged 65 years and over will increase from 323 million as of 1999 to 1.56 billion by 2050; a nearly five fold increase [ 5 ]. Head Injuries in the Elderly Basic Anatomy of the Intracranial Cavity The brain is located in, and protected primarily by the skull. The bones of the skull involved in non facial impacts from falls include the occipital (back), temporal (lateral), parietal (superior la teral), and frontal bones. There also exists a secondary protection system for the brain that consists of three layers of fibrous tissue (the meninges) and cerebrospinal fluid (CSF). The toughest and outermost fibrous tissue layer is the dura mater, whic h is in closest proximity to the irregular boney structures of the interior of the skull. Beneath the dura mater is the arachnoid fibrous tissue layer, which is filled with CSF. Finally, the delicate pia mater lies beneath the arachnoid layer and comes i n direct contact with the outermost brain tissue. The CSF located in what is also known as the subarachnoidal space protects the brain by stabilizing it during movements of the head. This fluid layer resists the tendency of the brain to move relative to the skull during head movement via its incompressible behavior. In other words, the CSF works to keep the brain in the proper position within the skull [ 6 ]. Bridging veins span the "subdural space from the cerebral surface on their way to the venous sinu ses" [ 6 ].

PAGE 20

! ) Common Head Injuries Experienced During Falls Traumatic Brain Injury Traumatic brain injuries typically occur when the head comes into direct contact with another object, which imparts forces and accelerations to the head. Focal injuries can occur when the brain accelerates relative to the skull as a result of the imparted energy to the head. This action can cause a variety of contusions (bruising) to the brain tissue due to the brain coming into direct contact with the skull by impact ( coup or contrecoup contusions), or by sliding across the irregular boney structures of the interior surface of the skull (gliding contusions). More precisely, gliding contusions are associated with the increase in normal and shear stresses at the interfac e of the skull and brain caused by the relative motion of the two, and the presence of trabeculae in the subarachnoidal space [ 6 ]. Lacerations are characterized by a tear of the brain tissue and are sometimes found near contusions, but typically arise fro m greater forces. Lacerations are more common in open (fractured skull) head injuries. Hematomas arise when bleeding results in the accumulation of blood in a closed space and can be classified as epidural, subdural, or subarachnoidal (intracerebral) dep ending on their location. Hematomas result from focal applications of force that tear arteries or veins [ 7 ]. When the CSF is unable to maintain the brain's position within the skull and relative motion between the skull and brain does occur, these veins are at high risk of rupture because they are attached to two surfaces moving in different directions [ 6 ]. Acute subdural hematomas (ASDH) are commonly caused by this mechanism and carry a particularly high rate of mortality at 74%, with little clinical tr eatment available [ 6 ]. There is potential, however, for TBI to arise without direct contact with another body and is instead due solely to acceleration of the head. In these types of injuries, the brain's inertia causes damaging strains throughout its ti ssue. One type of such injury is diffuse axonal injury (DAI). DAI is believed to result from angular acceleration of the head, which causes the outer tissue of the brain to move relative to the inner tissue resulting in widespread damaging tissue strains [ 7 ]. Other types of brain injury include brain swelling and increased intracerebral pressure. These injuries are caused by various mechanisms and tend to affect the brain in

PAGE 21

! a diffuse manner by limiting the blood supply to the brain (ischemic/hypoxic damage) [ 7 ]. Head injuries can be classified as either primary or secondary. Primary injuries are those that occur at the instant of injury (i.e. fall, impact, etc.), while secondary injuries evolve later on, often in response to primary injuries, and may complicate primary injuries. Examples of primary head injuries include skull fractures, contusions, lacerations, diffuse axonal injury (DAI), and intracranial hematomas. Secondary head injuries include brain swelling, hypoxia/ischemia damage, elevate d intracranial pressure, and even infection. One study of 635 fatal non missile head injuries focusing on the distribution of each of these types of injury found the following rates of occurrence: skull fracture 75%; DAI 29%, intracranial hematoma 60%, elevated intracranial pressure 75%, ischemic brain damage 55%, brain swelling 51%, and intracranial infection 4%. This study also found that males were much more prone to fatal head injuries (78%) than women (22%), and that falls were cited as the cause o f the injuries in 35% of the cases [ 7 ]. Another study of 263 co nsecutive head injured patients aged over 65 years [1] found that falls were the leading cause of injury (72%), followed by road accidents as pedestrians (19%). The types of injuries found in these cases were as follows: chronic subdural hematoma 29%, contusions 28%, acute subdural hematoma 21%, concussion 13%, DAI 7%, and acute epidural hematoma 3%. Of these injuries, 63% of acute epidural hematomas resulted in death, 27% of brain contusio ns resulted in death, 33% of acute subdural hematomas resulted in death, and just 3% of chronic subdural hematomas resulted in death. A study in Finland found that the rate of hospitalization for TBI was 18.1/100,000 for the entire population, but rose 59 .4% for patients aged 70 years and over [ 4 ]. The study found that the most common external cause of all TBI reported here were falls. Traumatic brain injury is known to lead to general spasticity, which is a major risk factor for subsequent falls [ 5 ]. O ther sequelae of brain injury include motor, sensory, cognitive and behavioral dysfunction, paresis, sensory disturbances in the feed forward or feed back mechanisms, decreased coordination, and motor control disorganization [ 8 ]; all of which may be risk f actors for falls.

PAGE 22

! + Skull Fracture The human skull is a highly irregular geometric shape, having widely varying curvature and thickness from one location to another. Furthermore, the skull is comprised of multiple bones whose cranial sutures represent di scontinuities in the material properties of either of the adjacent bones. Due to these complexities, the effective resistance of the skull to fracture varies greatly depending on the site of application and the orientation of an external force. When skul l fractures do occur, the resulting injuries are often severe. Penetrating head injuries run the risk of direct contact of the brain with the impacting object. Also, the fractured skull fragments can damage the underlying brain tissue by causing lacerati ons and bleeding. Risk Factors Blood Thinners A major factor that places older adults at greater risk of TBI is the use of aspirin which may affect the proper functioning of platelets, and anticoagulant therapies in the routine management of chronic co nditions like atrial fibrilation [ 4 ]. Anticoagulants inhibit blood from clotting, and in the case of a TBI this can have devastating consequences. Age R elated Factors The risk of falls among the elderly is known to increase with factors such as disabi lity of the lower extremiti es, musculoskeletal conditions, muscle weakness, and problems with balance, mobility, and gait [ 5 ]. It is also known that stroke patients are at an increased risk of falling due to "an upper motor neuron syndrome characterized b y spasticity, muscle weakness, and a variety of motor control abnormalities that impair the regulation of voluntary movement" [ 5 ]. Spasticity is thought to affect a patient's balance, mobility, and gait as it can cause exaggerated reflexes, rigidity, dyst onia, spasms, weakness, fatigue, and slow initiation of movement [ 5 ]. Also, co morbidities due to the

PAGE 23

! ! presence of two or more of any of these conditions is known to further increase one's risk of falling [ 9 ]. Rigidity and exaggerated reflexes were found in a study [ 10 ] that sought to quantify the differences in bracing for falls between the elderly population and a younger, university aged population. The testing found that the rigidity and over reaction of the upper extremity during falls among the olde r adults caused 10 15 times greater peak joint forces, making fractures of the upper extremity much more likely [ 10 ]. In the case of a fracture of the upper extremity during a fall, one's ability to further protect one's self against the remaining impact with the ground may be significantly hindered. This may increase the chances of a head impact with the ground. "Other normal aging changes include cerebrovascular atherosclerosis and decreased free radical clearance. The former could increase the risk o f injury or cause a secondary insult, and the latter may increase oxidative damage after TBI" [ 4 ]. Bone mass and strength is also known to decrease with increased age (after some maximum point attained around the age of 30) [ 11 ], which can make one's bone s more susceptible to fractures, including those of the skull. The effects of loss of bone mass can be especially problematic for women who become osteoporotic later in life. Osteoporosis can result in the loss of as much as 20% of peak cortical bone mas s and 50% of peak trabecular bone mass. The associated overall strength degradation of osteoporotic bones results in fragility and increased risk of fracture [ 11 ]. As the human brain ages, it becomes less elastic and its veins and arteries become more fr agile [1]. These changes in mechanical properties work to increase the risk of vein or artery tear and brain tissue laceration/contusion during an impact event. Muscle Strength Degradation Skeletal muscle strength is known to peak during one's twentie s, decline slowly until the age of about 50, decrease more rapidly afterwards, with greatly accelerated degradation after the age of 65 [ 11 ]. This decline is due to the loss of muscle mass, muscle fibre size, number of fibres, and quality of muscle. Ther e is a greater loss of fast twitch muscle fibres than slow twitch muscle fibres, which leads to a disproportionate

PAGE 24

! ! loss of strength with associated loss of muscle mass since the fast twitch fibres generate greater force [ 11 ]. Evidence also exists that s hows older muscles take longer to contract upon excitation and longer to relax after excitation. This is due to the degradation of motor unit firing rate and synchronization [ 11 ]. Also, older joints are usually not capable of producing forces to their ma ximum potential because of "antagonist co activation". This is when the muscle(s) working to extend a joint is being partially countered by the muscle(s) working to flex that joint. It occurs in older adults as a way of protecting the joint, but also res ults in lesser resultant forces being generated about the joint [ 11 ]. These decreases in skeletal muscle strength and force production about joints can lead to decreased ability to save oneself from a slip, trip, or fall. Brain Atrophy As the human bra in ages it may begin to atrophy, "which can cause occult findings to be present on head computed tomography (CT) despite an initial intact neurological examination" [ 4 ]. These changes may contribute to the development of gait dysfunction and impaired prot ective postural reflexes. Prognosis of Fall Related Head Injuries in the Elderly Older age is known to negatively affect recovery outcomes following TBI [ 12 ]. "The age adjusted rate of hospitalization for nonfatal TBI in the general population is 60.6 per 100,000 population; for adults aged 65 and older, this rate more than doubles to 155.9" [ 12 ]. Another study found this rate to be 203.9 per 100,000 for adults aged 75 and older, which is more than three times that f or the general population [12]. Th ese numbers show that hospitalization following nonfatal TBI is disproportionately high in the elderly population. Social and Economic Costs Emotional disturbance is a common social cost experienced after an acquired brain injury and has been shown to ad versely affect recovery, rehabilitation, and reintegration into the community [ 13 ]. "Changes in behavioral and emotional

PAGE 25

! %. functioning include depressed mood state, frustration and anger outbursts, reduced tolerance to stress, a diminished ability to show a nd control emotions, lack of initiation and reduced spontaneity" [ 13 ]. Furthermore, as an acquired brain injury patient begins to realize post injury changes to their life and their affect on the patient's future plans and goals, emotional distress and de pression increase [ 13 ]. One study of elderly community members found that nearly 50% of those who fell were afraid to fall again, and as a result 26% of those limited their normal activities [ 5 ]. In persons aged over 65, TBI results in over 80,000 emer gency department visits per year, with 75% of these injures resulting in hospitalization. In 2003, the total costs associated with the diagnosis of TBI in persons aged 65 and older was greater than $2.2 billion [ 12 ]. These tremendous monetary costs coupl ed with the rapid growth of this population will surely result in future challenges unlike any the healthcare industry has yet seen. Another study attributed these additional health care costs to the extended length of stays, additional diagnostic procedu res, surgeries, and possible litigation following fall events [ 14 ].

PAGE 26

! %% HELMET TESTING STANDARDS Anthropomorphic Test Dummy (ATD) Headform The current study utilizes the Hybrid III ATD headform manufactured by Denton ATD, Inc. which repr esents a 50th percentile adult male. Cadaver testing has been employed extensively in an effort to quantify various thresholds of the head to injury [ 15 16 17 18 19 20 ]. Generally in cadaver tests, accelerometers are securely fastened to the outsi de of the head via bone screws, glue, or a combination. One of the main issues with this approach, however, is that the center of gravity of the cadaver head cannot be accurately determined. Without precise knowledge of the location of the center of grav ity of the head, the accelerations measured at the outside of the head cannot be used to calculate the acceleration at the center of gravity of the head, from where all injury predictors take their acceleration values. Furthermore, the center of gravity o f the cadaver head is likely to move during impact as deformation occurs. These issues will result in injury criteria of limited confidence. The center of gravity of the ATD headform, however, is precisely known and does not move significantly during the types of impacts with which this study is concerned [ 15 ]. However, as noted in cadaver testing, human heads actually do deform during impacts while the ATD headform does not, highlighting a key limitation to the use of ATD headforms. Bio fidelity The Hybrid III ATD headform developed by Dr. Robert P. Hubbard and Donald G. McLeod is the most widely used artificial head in impact testing events, especially those involving automobile crashes. As outlined by McHenry [18] :

PAGE 27

! %& "Important features of the Hy brid III head are: Size, shape, weight and mass moment of inertia about the lateral axis that passes through its center of gravity that are representative of a 50th percentile adult male Anatomically correct location of its center of gravity and head to ne ck interface Humanlike forehead impact response Excellent impact response repeatability and durability for cranial impacts Excellent mid sagittal plane symmetry of response for cranial impacts Excellent reproducibility of response between heads" Head Prot ection Measures Many head injury predictors have been suggested and studied over the past few decades to determine which most accurately and consistently predicts head injury in humans. These include the Head Injury Criterion (HIC), peak resultant linear acceleration of the center of gravity (COG) of the head, peak resultant rotational acceleration of the COG of the head, linear impact velocity, angular impact velocity, Generalized Acceleration Model for Brain Injury Threshold (GAMBIT), Head Impact Power (HIP), peak force, time duration limits of several of the above, and much work in finite element analysis (FEA) and multi body dynamic software. The efficacy of the materials and material combinations tested will be quantified by their ability to limit or reduce the following head protection measures as found in the literature. Head Injury Criterion Currently, the most widely accepted predictor of head injury is the Head Injury Criterion. The HIC was developed following the Wayne State Tolerance Curve ( WSTC), which plots resultant translational acceleration (in multiples of E arth's gravitational constant, g which is equal to 9.81 m/s 2 ) of a cadaver head versus time duration at that acceleration. The resultant translational acceleration of the head is f ound by taking the root mean square (RMS) of the three orthogonal axes of acceleration data, as shown below:

PAGE 28

! %' Equation 1 : Root mean square of acceleration data where: RMS denotes root mean square AP denotes the Anterior Po sterior axis ML denotes the Medial Lateral axis SI denotes the Superior Inferior Axis Impact events falling above the WSTC are considered "dangerous to life", while those falling below the curve are not. See figure 1 below. Figure 1 : Wayne State Tolerance Curve (WSTC) The HIC is a mathematical model that closely follows this curve and whose value predicts head injury severity. The formula for HIC is given below:

PAGE 29

! %( Equation 2 : Head Injury Criter ion (HIC) formula where: t 1 and t 2 are any two times during the acceleration time history a RMS (t) = resultant translational acceleration of the head in G's t 1 t 2 are selected so as to maximize HIC Further developments of the HIC value included th e differentiation of HIC into the HIC 15 and the HIC 36 The HIC 15 sets an upper limit of 15 milliseconds on the value of (t 2 t 1 ), while the HIC 36 sets an upper limit of 36 milliseconds on the value of (t 2 t 1 ). As of 2000, the National Highway Traffic Safe ty Administration ( NHTSA ) officially adopted the HIC 15 over the HIC 36 as the more reliable predictor of head injury. Also, short duration impact event injuries like those being studied here are more conservatively predicted using the HIC 15 rather than the HIC 36 The HIC 15 predicts an impact event to be "dangerous to life" for any value grea ter than or equal to 700 [18]. One study found that a HIC value of 1000 corresponds to a probability of death of 10%, while a value of 2000 corresponds to a probabilit y of death of 50%. Another study found that a HIC of 1000 corresponds to an Abbreviated Injury Scale (AIS) of 2 (moderate), while a value of 1500 corresponds to AIS of 3 (serious) [ 21 ]. The COST 327 Report on Motorcycle Safety Helmets found that of all t he parameters analyzed, the HIC provided the best head injury severity prediction with a correlation coefficient of r = 0.80 [ 21 ]. The ability of the HIC value to predict skull fracture is poorly understood. One study found that the smallest HIC value as sociated with a fracture was 450, while the largest HIC value associated with a non fracture was 2351 [ 20 ]. This demonstrates that another predictor should be employed for the probability of skull fracture.

PAGE 30

! %) Peak Force Skull fracture has been shown to co rrelate more strongly with peak force than with pulse duration, loading rate, or resultant translational acceleration of the head. Using a one square inch impacting striker head, multiple pendulum impact tests were run on cadaver specimens of the age rang e we are focusing on [22, 23] "Clinically significant fractures" resulted from exceeding the following peak force thresholds at the indicated impact site of the skull: 4,003 N ( 900 lbf ) at the frontal area and 2,002 N ( 450 lbf ) at the temporoparietal are a. A "clinically significant fracture" is one that is readily detectible by palpation, x ray, anatomic dissection, or the use of dye penetrants among other techniques. Fractures of this severity put the brain at risk of further damage in addition to that experienced from the acceleration of the initial impact. Head drop tests exceeding these thresholds will be deemed unsuitable for this application. One reference [ 24 ] determined that skull fracture occurs on average at a kinetic energy at impact of ab out 68 N m ( 50 ft lbs ) onto a flat, unyielding surface. From simple physics this kinetic energy would be attained by dropping a 50th percentile male headform from a height of 1.2 m ( 4' 1 1/2" ) This reasoning neglects the fact that threshold of the skull to fracture is highly variable depending on location and orientation of the impacting blow. Also, the surface area over which the blow is suffered is not specified. This is problematic because a given kinetic energy transferred across a small area versu s a large area will have very different associated peak forces. Other studies aimed at determining a threshold for peak force of the skull found that skull fractures occur anywhere from 4,000 15,000 N (900 3372 lbf) [ 15 ]. One study of 31 cadaver heads being dropped onto a flat surface found that the average force required for fracture was 12,400 N (2,788 lbf) [ 16 ]. It has also been shown that the skull's peak force threshold decreases with decreased surface area of the impacting force [ 15 ], suggesting that fracture is more a function of stress (force/area) than force alone [ 16 ]. It has also been shown that fracture thresholds tend to be higher for males than females, and for frontal impacts than lateral impacts [ 15 ]. Finally, it should be noted that peak acceleration and peak force are near perfect correlates [ 19 ]. This is due to the fact that force is equal to mass multiplied by

PAGE 31

! %* acceleration according to Newton's Second Law, and mass is a constant term (mass of the headform) in the current study. Peak Resultant Translational Acceleration In addition to the HIC value and peak force limitations, it has been shown that peak resultant translational acceleration of the head exceeding 300 g is "dangerous to life" regardless of time duration of that acceleration event [25] Materials tested will therefore be deemed inappropriate if accelerations above this value are measured. Newman et al. suggested that resultant linear accelerations of the h ead from 200 250 g correspond to an AIS of 4 (severe), 25 0 300 g correspond to an AIS of 5 (critical), and those greater than 300 g correspond to an AIS of 6 (fatal) [ 21 ]. Peak Rotational Acceleration Rotational acceleration of the head has been used as a predictor of brain injury but is still poorly understo od. One study found that rotational accelerations of 4,500 rads/sec 2 proved fatal in some tests, while other experiments found that 16,000 rads/sec 2 proved non injurious. In the current study, the rotational behavior of the headform during/after impact w hile detached from the rest of the dummy is trivial and should be neglected. The COST 327 Report on Motorcycle Safety Helmets found that resultant linear accelerations and resultant angular accelerations had a large linear correlation coefficient (r = 0.9 3) when the headform was both attached to the rest of the dummy and detached from the dummy. From this knowledge it may be reasonably deduced that reducing linear accelerations in our tests would reduce angular accelerations in a roughly linear fashion. Gadd Severity Index The Gadd Severity Index (GSI) was developed by C. W. Gadd in 1966 as a mathematical best fit to the head injury data available to him at the time. The injury prediction model he produced is given as follows:

PAGE 32

! %+ Equation 3 : Gadd Severity Index (GSI) formula where: t 1 and t 2 are any two times during the acceleration time history a RMS (t) = resultant translational acceleration of the head in G's t 1 t 2 are the beginning and ending times of the impulse A GSI value greater than or equal to 1000 was considered unacceptable [ 25 ]. It is clear from the above equation that the Head Injury Criterion used the GSI as a starting point for its refinement. Though the GSI was developed first, the HIC is more widely accepted today Headform Drop Testing Methods The drop test design is an important step in the evaluation of head protection devices, as some tests may affect head protection measures in an inconsistent or otherwise undesirable manner. Careful considera tion has been given to the following test designs and the design thought best for this application was selected. Guided Drop Tests Guided head drop test setups are advantageous as they allow for highly repeatable test results. This allows the experiment er to run multiple tests for a given scenario, thereby increasing the reliability of measured results and providing a means of statistical analysis. Also, since the headform is impacting in nearly the exact same orientation every trial, one can use a sing le axis accelerometer as opposed to a tri axial accelerometer to capture the data of interest as long as the accelerometer is mounted normal to the impacted surface and located at the center of mass of the headform. Twin Wire Test Apparatus The twin wi re apparatus is one such guided drop test approach. The headform is rigidly mounted to a drop carriage, which is guided by two steel wires from the drop height all the way to the impact surface.

PAGE 33

! %, Monorail Test Apparatus The monorail design is similar to that of the twin wire, except it employs a monorail guide track, which the drop carriage runs down, instead of steel wires. The drop carriage includes low friction wheels that run along the monorail track. Limitations Drawbacks to the guided fall desig ns occur during the impact event. The headform impacts with the momentum of itself plus the drop carriage, introducing undesirable resonances into the acceleration data. Also, instead of being allowed to rebound off the impacted surface naturally, the hea dform is limited by the drop carriage, which is held in place by either the two wires or the monorail track. Freefall Drop Tests Platform Test Apparatus The platform drop test starts with the headform on a platform at the desired drop height. The entir e platform is then allowed to drop along guide wires onto the impacting surface below. The headform is loosely kept on the platform by a net, which imposes no direct interference on the headform's behavior during the initial impact and only serves to keep the head from rolling off the platform after impact. True Freefall Test Apparatus The true freefall test apparatus is one that seeks to overcome the drawbacks of the aforementioned approaches by allowing the headform to drop completely unconstrained o nto the impacting surface in the most repeatable way possible. This is done by placing the headform over a trap door located at the center of a height adjustable platform. The trap door is designed so as to minimize rotation imparted to the headform whil e it falls through.

PAGE 34

! %! Limitations The platform drop test has the drawback that the platform wire interface will certainly produce some amount of friction, thereby hindering the headform's natural descent. Also, the material properties of the platform (w hich the headform impacts directly) will almost certainly differ from those of the floor, affecting the impact characteristics that would normally be seen in a hospital setting. The true freefall apparatus is limited in that the rotation of the headform i s nonzero, therefore varying the impact orientation of the headform from one trial to the next. This makes statistical analysis not feasible. Comparison of Guided and Freefall Tests Both the guided and freefall drop test methods have certain unavoidab le disadvantages as well as unique advantages. The general limitation of the guided tests is their guidance system. Any form of guidance necessarily introduces friction to the test, which works to slow the descent of the headform. This friction may also be variant in nature, leading to a range of results at a given drop height. The drop carriage and guidance system may also introduce unknown acceleration characteristics that may not be representative of the headform's behavior. The platform testing met hod reduces the guidance rebound issues, but does little to overcome guidance frictional effects. The true freefall approach eliminates both rebound issues and guidance friction, but is not the most highly repeatable method. Considering the advantages an d disadvantages of the above approaches, this study will utilize the true freefall approach for it's true impact characteristics and will present a 95% confidence interval statistical approach to address repeatability issues in the results Various Indust ries Studied Since no head protection standards have been put in place for medical helmets, other industries were referenced for their head protection protocols. Below is a summary of what was found for each of the industries examined.

PAGE 35

! &. Medical Helmets The healthcare industry currently uses several protective medical helmets in an effort to protect those patients most prone to falls and those most at risk if a fall were to occur. As stated previously, no standards of protection have been established i n this field. Prescription of these helmets is not necessarily enough, however, since compliance remains a major barrier. The current bulky appearance of these helmets may be a contributing factor to non compliance as patients are unwilling to wear such conspicuous headwear. Dilatent and honeycomb materials are appealing in this regard as they could provide superior impact attenuation characteristics per unit thickness over the existing foam, plastic, cotton, and leather helmet construction materials. S uch dimensional advances may allow for a thinner, less obvious head protection device coupled with increased compliance and comparable, if not superior, protection performance. Motorcycle Helmet Standards One European study found that the risk of injury to motorcyclists was ten times that of automobile drivers [ 21 ]. While rates of wearing safety helmets are high, head injuries remain the leading cause of serious and fatal motorcycling injuries [ 21 ]. Because of this, much time and effort has been investe d into the causes of such head injuries and how motorcycle helmets can be improved to better protect against them. The result of this effort comes in the form of various helmet standards aimed at protecting against the types of forces and accelerations co mmonly seen in motorcycle accidents. The studies backing these standards deal with establishing thresholds on what the human head can sustain before injury occurs. Below is a summary of several of the more prominent standards and their injury thresholds. Snell Standard The Snell Memorial Foundation is a private, non profit organization that created and continuously updates the voluntary Snell M2005 standard. Snell testing uses a monorail or guide wire drop assembly to drop the helmeted headform onto flat, hemispherical, and sharp edge surfaces at speeds ranging from 6.6 7.8 m/s. A helmet

PAGE 36

! &% passes these drop tests only if the peak resultant linear acceleration of the headform is less than or equal to 290 g This standard involves very high energy i mpacts and has been accused of efficiently absorbing energy only during impacts that are not likely to be survivable in the first place [ 25,26 ]. In other words, this standard may produce a helmet that is too resilient, and not practical for the types of i mpacts a motorcyclist is most likely to face on the road. DOT Standard The Department of Transportation (DOT) in the U.S. created the FMVSS No. 218 standard (often referred to simply as the DOT standard) to establish a certain minimum level of protection in motorcycle helmets. This testing uses a monorail drop assembly to drop the helmeted headform onto flat and hemispherical surfaces at speeds ranging from 5.2 6.0 m/s. A helmet passes these drop tes ts only if the peak resultant linear acceleration of the headform is less than 400 g This standard further specifies that the resultant linear acceleration of the headform may not exceed 200 g for more than 2 milliseconds or 150 g for more than 4 milliseconds. This standard is unique in this regard, as i t is the only standard to address time duration limits on lesser accelerations. BSI Standard The British Standards Institution (BSI) created the BS 6658 standard for motorcycle helmets. This standard uses a form of "guided fall" (not further specified) to drop the helmeted headforms onto flat and hemispherical surfaces at speeds ranging from 4.3 7.5 m/s. A helmet passes these drop tests only if the peak resultant linear acceleration of the headform is less than or equal to 300 g ECE Standard The European Community standard No. 22.05 employs an unrestrained headform drop test methodology. The helmeted headform, with attached neck segment, is dropped onto a flat and curb surface at 7.5 m/s. A helmet passes these tests only if the peak

PAGE 37

! && resultant li near acceleration of the headform is less than or equal to 275 g and the HIC value is less than or equal to 2400. This methodology has been criticized because the unrestrained headform is free to rotate on impact, resulting in unmonitored rotational accel eration and possibly lessened peak linear acceleration. Soccer Helmet Standards One study [ 27 30 ] determined that risk of mild traumatic brain injuries (MTBI) in soccer are well predicted by peak linear acceleration and maximum "Head Impact Power" (HI P) and can be quantified according to the following: Table 1 : Risk levels with associated head injury measures Risk Level 5% 50% 95% Peak Linear Acceleration 40 g 78 g 115 g HIP max 4.5 12.8 21.3 Head Impact Power was developed in reference to professional American football players and is a function of linear and angular head accelerations and velocities. More specifically, HIP computes a time rate of energy transferred to the head. Again, since the current study doesn't measur e angular quantities this head protection measure will not be included. There is no established standard for head protection devices in soccer, nor a requirement that any player wear one. However, several head protection technologies aimed at protecting a player's head from injurious impacts are available on the market today. The forces and accelerations that these technologies aim to lessen are generally determined by, and therefore unique to, each individual manufacturer. Also, it has been noted that the clinical and biomechanical effectiveness of these technologies is still poorly understood [ 27, 28, 29 ]. Playground Surfacing Materials Standards The current standards for playground surfacing materials test playgrounds by dropping an instrumented headform from a height equal to the highest accessible height

PAGE 38

! &' on the playground equipment. To pass the test, the peak resultant linear acceleration must be less than or equal to 200 g and the HIC value must be less than or equal to 1000 [ 31 32 ]. Rugby H elmet Standards According to the International Rugby Board (IRB) standard, rugby helmets are dropped from a height of 30 cm using a twin wire guided drop assembly onto a flat, rigid surface. A helmet passes the test with a peak resultant linear accelerat ion of no more than 550 g but no less than 200 g and a Gadd Severity Index (GSI) of no more than 1200. This minimum peak acceleration requirement is unique to this industry and included because "headgear that perform under this threshold could cause pla yers to use their heads more, risking cervical spine injury", according to the IRB [ 33 ]. Bicycle Helmet Standards One standard for the design of bicycle helmets limits the peak force transferred to the head to no more than 10,000 N (2,248 lbf) with the a dditional requirement that the foam liner not "bottom out", or crush completely [ 34 ]. This stipulation is likely due to the fact that a bottomed out liner transfers nearly all of the additional energy from the outer shell directly to the head, attenuating a negligible amount. The reason foam liners work is because the action of crushing the foam requires energy. Therefore, the energy transferred to the head will be less than the amount transferred to the outer shell by an amount equal to the energy spent crushing the foam liner. Conclusion Having studied many industries involved in head protection and quantification of head protection measures, experimental methods can now be designed. This study will focus on peak acceleration, peak force, and HIC15 a s the measures of head protection, and the tests will be carried out using the true freefall head drop test method. Since these head protection measures are not exact or certain predictors of head injury, any comparisons made about protection will be rela tive rather than absolute.

PAGE 39

! &( METHODS Introduction The testing in this study is divided into two phases. Phase I was aimed at assessing the impact attenuation characteristics of individual materials as well as materials in combination. However, befo re Phase I could be carried out, some preliminary tests were required. Phase II was aimed at assessing the impact attenuation characteristics of existing medical helmets and three prototype designs, which incorporate those materials tested in Phase I. Th e results of Phase I lead to the design of the prototypes tested in Phase II. Due to the serial relationship of the testing, preliminary testing methods will be presented, followed by Phase I and Phase II testing methods. Location The current study wa s conducted at the James A. Haley Veterans Administration Patient Safety Center of Inquiry Biomechanics Laboratory in Tampa, Florida. Preliminary Head Drop Testing Apparatus Design and Construction The headform used in this study is that from the Hybrid III anthropomorphic test dummy (ATD) designed and manufactured by Denton ATD, Inc. The headform size and mass are representative of a 50 th percentile male. The bio fidelity of the headform has been we ll established in past studies [18] to assure accurat e size, shape, and mass of the representative population. The headform has a tri axial accelerometer located at its center of mass. The three orthogonal accelerometer axes are anterior posterior (AP), superior inferior (SI), and medial lateral (ML). Thi s setup allows for acceleration measurement in each axis throughout the impact event. Taking the root mean square (RMS) of these three orthogonal acceleration components gives the resultant acceleration

PAGE 40

! &) of the center of mass of the headform at all times d uring the impact event. The resultant acceleration of the headform's center of mass, reported in multiples of the earth's gravitational acceleration ( g = 9.81 m/s 2 ), is a widely accepted convention for translational head acceleration magnitude as an injur y criterion. With this acceleration time history one can find peak acceleration values as well as HIC values. The drop tower (also referred to as the "gallows") was designed and constructed by John Lloyd and Shawn Applegarth. The frame was constructed of steel Uni Strut members and was mounted on four lockable wheels for easy mobility between testing sessions. A Bianca bed lift was mounted at the top of the drop tower. A smaller, separate frame was constructed from steel angles and suspended from the Bi anca bed lift such that it could easily be raised or lowered to the desired drop height within the larger drop tower frame. This smaller height adjustable frame ran on track in the main tower's frame for increased stability and proper alignment. The heig ht adjustable frame supported a plywood platform having a trap door at its center. The headform was placed directly onto the center of this trap door. Striking a large trigger pin with a rubber mallet activated the trap door. This pin in turn activated a spring loaded doorstop, which quickly released the trap door's support from below, allowing the doors to quickly fall out from under the headform. The trap door was designed to minimize unwanted rotation imparted to the headform by the trap doors at the instance of drop. To this end, the trap doors were covered with a thin polymer sheet having a very low coefficient of friction. This addition was motivated by review of early drop tests and noting some unwanted rotation during the free fall portion of t he drop. This rotation is to be avoided, as it introduces repeatability issues to the test method. Another addition to the apparatus included springs attached to the bottom of the trap doors. These springs connected the trap doors to the height adjustab le frame and worked to pull the trap doors down faster than would be seen by gravity working alone. Again, this addition was intended to minimize drag between the headform and the trap doors, thereby minimizing unwanted rotation. Eliminating headform rot ation allowed for more repeatable impact location on the headform. The headform was dropped onto a force plate designed and constructed by John Lloyd and Shawn Applegarth. The force plate initially consisted of two square shaped

PAGE 41

! &* 6.35 mm ( ") thick stee l plates each measuring 762 mm x 762 m m (30"x30"). The two plates were sep arated vertically by a gap of 19 m m ( "). This gap was maintained by four piezoelectric load cells (or force sensors); one located at each of the four corners of the plates. Upon review of early head drop tests recorded using a high speed camera, it was noted that the top plate deflected considerably approximately 6.35mm ( ") upon impact, which may have been effectively reducing the peak force and peak acceleration measured by increasing the stopping distance of the headform. This concern was remedied by modifying the top plate by decreasing its area considerably to 40.64 cm x 40.64 cm (16"x16") and further reducing the footprint of the force sensors to 20.32 cm x 20.32 cm (8"x 8"). This modification theoretically reduces the top plate's deflection from about 6.35 mm ( ") to about 0. 4 mm ( 1/64"). Another concern with the original force plate was that the entire instrument was "hopping" off of the floor after impact. This was r emedied by screwing the bottom plate directly into the concrete floor. Further testing showed much improvement in both areas of concern. The force sensors (model 208C05) were designed and manufactured by PCB Piezotronics. The sensors have an operating r ange of approximately 0 to 22,250 N (or 0 to 5,000 lbf) and output a corresponding voltage signal between 0 and 5V. The force sensors use quartz crystals for accurately measuring dynamic forces and consequently have an associated time constant effect for static loads. This means that attaching the heavy steel plate on top of the sensors is acceptable, as the voltage signal will approach zero with first order behavior and a time constant of approximately 2,000 seconds. In other words, the force exerted on the supporting force sensors by the top steel plate will no longer be observed after approximately 35 minutes, and the force sensors will only respond to additional dynamic loading thereafter.

PAGE 42

! &+ ! Figure 2 : Head drop testing to wer Protocol The preliminary head drop tests were carried out at 1.0, 1.2, 1.4, 1.6, and 1.8 meters. The one meter drop represents a moderate fall, while the two meter drop represents a worst case fall from a standing position for a 1.8 meter tall perso n, or a fall from a person seated in a high bed. The headform was dropped onto the force plate once from each of the above listed heights with no head protection. This provided some baseline data, which was used in designing the impact testing protocol ( i.e. impacting force), as well as debugging the MATLAB code. During these initial tests, the two second sampling window was initiated by the data collector stationed at the PC via the "start data collection" button in the Vicon Workstation interface. The data collector called out a countdown to data collection initiation so that the assistant striking the trigger pin could sync their drop initiation. From these initial tests it was determined that the unprotected head experienced a peak force of 15,000 N at the worst case scenario. This peak force was therefore applied to the materials tested in Phase I using the impact tower.

PAGE 43

! &, Data Collection, Calibration, and Verification The force sensors needed to be individually calibrated to determine the conversi on factors for sensor output to Newtons. This would be accomplished using a handheld Chatillon gauge, which measures tensile and/or compressive forces. Before this could be used, however, its own accuracy had to be verified through a series of tension ca libration tests. This was simply done by hanging various known weights from the gauge's hook and confirming the known input weight (lbf) to force output (N) relationship. An excel chart was employed to determine the best fit line passing through the orig in for the data obtained. The slope of this best fit line was found to be 4.401 N/lbf with R 2 =0.9997 (see figure 3) as compared to the theoretical conversion f actor of 4.448 N/lbf This discrepancy of 1.057% is deemed acceptable for the current study, a s the helmet will be designed to attenuate the maximum force by a much larger percentage. With the Chatillon gauge's accuracy verified, the force sensor calibration could be carried out with confidence. Figure 3 : Verification of Chatillon gauge accuracy

PAGE 44

! &! Figure 4 : Verification technique for Chatillon gauge The Chatillon gauge was then set to record only maximum compressive force during each compression trial. With the top plate removed from the force plate, the Chatillon gauge was pushed directly downward onto the force sensors in a systematic way. Each sensor receive d five successively larger compressions during which the maximum Chatillon readout was manually recorded while the force sensor ou tput was simultaneously and continuously recorded in the Vicon Workstation on a nearby PC. The five maximum force sensor outputs were then plotted against the five maximum Chatillon readouts in an Excel chart. A "best fit" line passing through the origin was then added to the plot. The slope of this line represents the conversion factor from force sensor output to Newtons for each sensor. An example of this chart is shown in figure 5 for force sensor "A". The complete results of this calibration method are shown below in Table 2.

PAGE 45

! '. ! Figure 5 : Example of conversion factor determination (force sensor A) Figure 6 : Compression test with Chatillon gauge on force sensor Table 2 : Determined conversion factors for indicated force sensors Force Sensor Designation F_A F_B F_C F_D Conversion Factor (output/N) 0.1464 0.15 0.1456 0.1414

PAGE 46

! '% Each of the four force sensors sends its voltage signal directly to the PCB signal conditioner (mod el 482A22) via four cables. After conditioning, the four analog signals from the force sensors are joined by the three channels (one for each axis) of accelerometer data and sent to the Vicon analog to digital converter (ADC) where they are converted to d igital signals and then sent to the Vicon Workstation software on the PC (see schematic below in figure 7 ). The Vicon Workstation interface is set to sample the signals at 12,000 Hz for duration of 2 seconds, resulting in 24,000 samples per channel per dr op test, or 168,000 samples per drop test considering all 7 channels. The collected data sets are then exported to Excel files. These Excel files are altered using a simple macro to remove column headings and then saved as text files (.txt), which are la ter referenced in the MatLab code for analysis. Figure 7 : Schematic of data collection

PAGE 47

! '& Phase I: Impact Testing Apparatus Design and Construction Impact testing of potential helmet materials was carried out using the INSTRON Dynatup (model 9250HV) tower. This tower operates by dropping a guided carriage with an impacting tup mounted beneath it onto the material of interest. The tup was secured to the drop carriage by a PCB force sensor (model 208A15) with capacity of approx imately 0 to 22,250 N (or 0 to 5,000 lbf). The impact tower was connected to a signal conditioner and PC for data collection and analysis through the Impulse data acquisition software package. The tup size and shape had to closely resemble that of the im pacting surface area of a human head during a fall in order to assure applicable results. To design an appropriate tup the surface area and shape of the site of the head that carries the impacting force must be determined. To do so, the 50 th percentile male headform was dropped from 1.5 meters into a shallow cardboard box (8.5"x10"x1.5"), which was uniformly filled with approximately 25 m m (or 1") of Play Doh. The headform was dropped onto different impact sites to observe the various impact site chara cteristics. The impacting sites tested were front, front/side, side, and back of the head. This approach provided clearly measurable surface area, shape, and curvature of the impacting site. Of the four impact sites, the front/side site was observed to have the deepest crater with smallest surface area (72 sq. cm). Since pressure is maximized when surface area is smallest (see equation 4) for a given force, this is therefore the worst case impact site and will be used for the tup design. Equation 4 : Pressure formula where: P = pressure (Pa) F = force (N) A = surface area (cm 2 )

PAGE 48

! '' The crater was roughly circular for every site test, so the tup will be circular in shape. A circle with a surface area of 72 square centimet ers has a radius of 48 mm. The front/side impact site crater was approximately two centimeters deep, so the tup was designed to be two centimeters thick. The curvature observed in the front/side impact site crater was uniformly curved throughout, and thu s looked like a portion of a sphere. The tup was designed to have this quality as well. The tup was fabricated of solid stock aluminum cylinder on a lathe with the help of Stuart Wilkinson. The radius and depth of the tup are easily measurable dimension s, while the curvature is not. Thus, the curvature of the tup was worked by hand using a metal file on the lathe and adjusted "by eye". From initial unprotected impact tests, it was seen that the two meter drop had the highest peak force of all drop heig hts and was approximately 15,000 N (or 3,372 lbf). This will be the impact force the impact tower will be set to mimic. It is important to understand that the impact tower cannot be set to deliver a particular peak force to the material since the recorde d peak force is inversely proportional to stopping distance. Therefore, the tup was dropped onto the cement block with no protective material present from increasing heights until the recorded peak force was within a couple of percentage points of the 15, 000 N goal. The height at which this occurred was about 7 cm (or 2.82" ) Thus, all materials will be tested from this height. With an appropriate impact height and impact tup, candidate helmet materials were tested using the impact tower to measure their impact attenuation characteristics. These impact attenuation properties were normalized so that all materials could be compared side by side. This normalization was done by dividing "percent peak force reduction" by "material thickness" to come up with the "effectiveness per millimeter". This gives the experimenter a way of determining if a material is effective because it is designed well, or simply because it is thicker than the other materials. The design of the material is the more significant of t he two qualities, as thickness can be adjusted based on this.

PAGE 49

! '( Protocol 5 dilatent materials available and 7 honeycomb materials were available for testing. These materials were each tested individually for five trials each (60 tests). Next, these mater ials were tested in every possible combination of one dilatent and one honeycomb material for five trials (175 tests). In the combination tests, the dilatent material was always placed on top of the honeycomb material. This was done because the shear thi ckening quality of the dilatent material may not be activated if it were placed underneath the honeycomb material, which would reduce the force transmitted to the dilatent material. Furthermore, the dilatent material is intended to stiffen upon loading in order to spread the load over a larger portion of the honeycomb material. This function could not be possible if the dilatent material were underneath the honeycomb material. Data Collection The data files generated from these tests were exported fr om the Impulse software package to Excel. In Excel, the digital data is converted to usable force data using appropriate conversion factors (provided in the Excel file by the Impulse software) and then the five tests per condition are compiled into one Ex cel spreadsheet. This Excel file is then saved as a text file for later reference in the MATLAB code. Data Processing The MATLAB code first imports the text files (.txt) generated in Excel. The raw data is filtered using a 4th order low pass Butterwo rth filter. Since five tests were run on each of the individual materials as well as for the materials in combination, the averages of these five trials were found resulting in one composite force time history trace for each material or material combinati on. Force vs. time and acceleration vs. time graphs can be derived by dividing force by the known mass of the drop carriage (provided in Excel files generated by Impulse software). Also, HIC values can be determined from the acceleration time histories These impact test results were used to narrow the field of potential materials.

PAGE 50

! ') Phase I I : Head Drop Testing Apparatus Design and Construction The apparatus used in the formal head drop testing phase was the same as that used in the preliminary stag es of head drop testing. Protocol A In order to gather more data on the materials that performed well in the impact tests, a new testing method was developed. The materials were placed directly onto the force plate and the unprotected headform was dro pped onto them. The top performing materials from P hase I included two honeycomb materials and four dilatent materials. These six materials were tested individually in this manner. The headform was dropped from an upside down orientation so that the top of the head would strike the materials. This was done because this orientation provided the highest level of repeatability in impact location. The headform was dropped three times from one meter in this orientation onto each of the individual materials. Next, the materials were tested in each possible combination of one honeycomb and one dilatent material, as well as each possible combination of two honeycombs and one dilatent material. Order of materials was also varied. This resulted in 35 scenarios including the individual materials testing. For this data, the ultimate and absolute performance rankings of the material combinations were determined by John Lloyd in order to determine the top performers from this set of testing. With the top three ma terial combinations determined, the three prototypes could be constructed. Protocol B The prototypes were assembled by cutting the top performing materials into approximately 4"x3" rectangles that could fit into the prototype shell designed and constru cted by Jim Ferguson. With the representative "stack" of materials inserted into the prototype shell each of the three prototypes were tested accordi ng to the protocol used during the preliminary head drop tests

PAGE 51

! '* The Plum and HipSaver medical helmets we re also tested on the drop tower These tests were carried out by dropping each helmeted headform from the five drop heights previously listed three times at each of the three orientations (front, back, and side). This means there are 45 dro p tests for e ach of the helmets and prototypes. The unprotected case was also re tested with the modified force plate. Data Collection As was done in the preliminary stages of head drop testing, the accelerometer in the headform and the force plate both send digital data signals to the nearby PC. These data files are partially modified in Excel, then formally analyzed using MATLAB code. Data Processing The MATLAB code imports the raw force sensor data and the raw accelerometer data. Variables are defined from the text file. The acceleration data is already in g 's (multiples of Earth's gravitational constant), but the force sensor data needs to be converted to Newtons using the aforementioned conversion factors. These data sets are next filtered using a 4 th order low pass Butterworth filter with cutoff frequencies for the accelerometer and force sensors of 1650 Hz and 1800 Hz, respectively. The cutoff frequency used for the headform accelerometer is widely accepted for this application [ 20 ]. The cutoff frequency used for the force sensors was found by taking the "Fast Fourier Transformation" (FFT) of the raw data and observing where the majority of the data existed and where higher harmonics existed. This cutoff frequency was chosen to retain the data believed to be representative of the impact event and to attenuate the data believed to be higher harmonic noise. At this point the four force sensor channels are summed to obtain the resultant force at each sample and the RMS is taken for the three axes of accelero meter data to obtain the resultant linear acceleration of the center of mass of the headform at each sample. The MatLab code next plots resultant forces and resultant accelerations versus samples, finds peak resultant force, peak resultant translational a cceleration, and finall y calculates the HIC15 value

PAGE 52

! '+ RESULTS Phase I Results Phase I results are shown in figures 8 and 9 below. These graphs (generated in MATLAB) show the acceleration of the impacting tup during the impact event with the indicated materials. The tabulated results from the MATLAB code can be found in appendix 2. Photographs of the indicated materials are presented in appendix 3.

PAGE 53

! "# Figure 8 : Acceleration vs. time traces for dilatent materials tested on the impacting tower.

PAGE 54

! "# Figure 9 : Acceleration vs. time traces for honeycomb materials tested on the impacting tower.

PAGE 55

! "# Phase II Results The results from dropping the unprotected head onto the force plate covered in various materials are presented in table 3 below. The top three performing materials/material combinations from this set of testing were used to construct the three prototypes for the final step of head drop testing. Note that the naming convention d escribes the order of materials as stacked from the force plate surface upwards vertically towards the impacting headform.

PAGE 56

! "# Table 3 : Results of head drop testing of indi vidual materials and materials in combination

PAGE 57

! "# Table 3 (Continued)

PAGE 58

! "# Below are presented the complete results for the unprotected head d rop tests with best fit lines. Note the irregular, nonlinear trends in all measures, with correspondingly low R 2 values. R 2 values are presented in the same order as the orientations in the legend; in other words the first R 2 value corresponds to the sid e orientation, the second to front, and the third to back.

PAGE 59

! "" Figure 10 : Head drop testing results for the unprotected case (peak acceleration).

PAGE 60

! "# Figure 11 : Head drop testing results for the unprotected case (HIC15).

PAGE 61

! "# Fi gure 12 : Head drop testing results for the unprotected case (peak force).

PAGE 62

! "# Next is presented a complete comparison of the three candidate prototypes, named "Prototype 1", "Prototype 2", an d "Prototype 3". Each of these prototypes consists of the same three materials (Hy, Hz, and Dd) stacked in a different order. A new naming convention was used for this phase of testing so that honeycomb and dilatent materials could be easily differentiat ed, therefore Hy was previously referred to as "D", Hz was previously referred to as "F25", and Dd was previously referred to as "S2". The two honeycomb materials are polymer materials fabricated to have a three dimensional honeycomb structure. These hon eycomb materials have unique impact attenuating characteristics (see figure 9) and work together to protect against a wide range of impact forces. The dilatent material is thin, white, and relatively stiff. These material combinations were determined from the head drop testing of individual materials and materials in combination. "Prototype 1" consists of HyDdHz, "Prototype 2" consists of HzHyDd, and "Prototype 3" consists of DdHyHz. HIC15, peak acceleration, and peak force are plotted against indicated drop heights for each of the three prototypes in the front orientation. Results for the other orientations can be found in appendix 2 but are not presented here because the front orientation is that which the headform was designed. 95% confidence interva ls are presented in order to draw statistically significant conclusions about the performance of the various prototypes.

PAGE 63

! "# Figure 13 : Prototype comparison (peak acceleration).

PAGE 64

! "# Figure 14 : Prototype comparison (peak force).

PAGE 65

! "# Figur e 15 : Prototype comparison (HIC15).

PAGE 66

! "# In order to draw a proper conclusion about which prototype performed best, 95% confidence intervals are presented in the preceding plots. Furthermore, these confidence intervals are presented numerically in the following tables and those cases that are statistically sign ificantly better than others are highlighted. Note the small number of cases exhibiting statistical significance. Table 4 : Results of prototype comparison (back orientation)

PAGE 67

! "# Table 5 : Results of prototype comparison (front orientation)

PAGE 68

! "# Table 6 : Results of prototype comparison (side orientation)

PAGE 69

! "# The best performing prototype ("Prototype 1") is next plotted on the sa me graphs as the Plum "ProtectaCap+Plus" medical helmet and the HipSaver medical helmet for direct comparison to current market technologies. Again, the same head protection measures are used for comparison and the drop orientations and heights are indica ted.

PAGE 70

! "" Figure 16 : Comparison of prototype, Plum, and HipSaver helmets (peak acceleration in the front orientation).

PAGE 71

! "# Figure 17 : Comparison of prototype Plum, and HipSaver helmets (peak force in the front orientation).

PAGE 72

! "# Figure 18 : Comparison of prototype, Plum, and HipSaver helmets (HIC15 in the front orientation).

PAGE 73

! "# As was done previously, the various head protection measures are presented with their associated 95% confidence intervals to show statistical significance of performance. The tables that follow highlight those cases where statistical significance exists between the various helmets with respect to the 95% confidence interval. Table 7 : Results of helmet comparison (b ack orientation)

PAGE 74

! "# Table 8 : Results of helmet comparison (front orientation)

PAGE 75

! "# Table 9 : Results of helmet comparison (side orientation)

PAGE 76

! "# Finally, the top performing prototype, "Prototype 1" (HyDdHz), is plotted along side the data from the unprotected case. This provides a direct comparison of how effective the prototype is against head injuries due to falls. Note that the unprotected data was gathered at height increments of 0.1 meter, while the prototype data was collected at 0.2 meter increments. While these extra drop height measures cannot be compared directly against the prototype at thes e heights, the data is presented for completeness. Again, 95%confidence intervals are presented in order to highlight statistical significance of results.

PAGE 77

! "# Figure 19 : Comparison of prototype 1 and unprote cted case (HIC15 in the front orientation).

PAGE 78

! "# Figure 20 : Comparison of prototype 1 and the unprotected case (peak acceleration in the front orientation).

PAGE 79

! "# As was done previously, the various head protection measures are presented with their associated 95% confidence intervals to show statistical significance of performance. The tables that follow highlight those cases where statistical significance exists between the prototype and the unprotected case with respect to the 95% confidence interval. Note the significance of improvement of the prototype over the unprotected case in almost every case. Table 10 : Results of comparison of "Prototype 1" and the unprotected case (back orientation)

PAGE 80

! "# Table 11 : Results of comparison of prototype and the unprotected case (front orientation)

PAGE 81

! "" Table 12 : Results of comparison of prototype and the unprotected case (side orientation)

PAGE 82

! "# DISCUSSION Interpretation From the impact testing of materials, it is seen that each material has a unique way of attenuating the impacting energy. The characteristics of these attenuation profiles are seen in the shapes of th e acceleration time history curves. Generally speaking, the materials that attenuate the impact energy quickly have higher peak acceleration values, while those that attenuate the energy more gradually have lower peak acceleration values. Based on the head protection measures HIC15, peak acceleration, and peak force found from the literature review, it can be seen in figures 16 18 that "Prototype 1" performed better than the other two prototypes tested. The order in which the materials are stacked was shown to affect performance. And it is seen that a dilatent material sandwiched between two honeycomb materials is best when using a three material combination. This order of materials is interesting since the dilatent material is designed to stiffen upo n impact. That functional design would suggest that the dilatent material should be placed on the outermost layer in order to shunt the impact force over a greater area of honeycomb materials underneath. The results, however, show that having a honeycomb material in the outermost layer attenuates some of the impact force, transfers the rest to the dilatent material, which then stiffens and distributes the remaining force over a larger area of the innermost honeycomb material. Comparing the top performing prototype's data to that of the Plum ProtectaCap+Plus and HipSaver medical helmets, figures 22 24 show that "Prototype 1" was again the top performer in the majority of the cases. However, the Plum ProtectaCap+Plus medical helmet is a close runner up. Figures 27 29 show the effectiveness of the top performing prototype as compared to the unprotected case. It can be seen that the prototype afforded statistically significant improvement over the unprotected case in all but one instance.

PAGE 83

! "# Limitations The current study employed a free fall head drop testing method in order to overcome the degree of freedom issues associated with the guided drop methods. While this aim was satisfactorily accomplished with the free fall method, repeatability issues were int roduced. Also, side orientation impacts tended to have less severe head protection measures. This is believed to be because the headform is impacting the force plate and immediately spinning laterally off to the side. This converts the linear kinetic en ergy of the headform into rotational kinetic energy, minimizing the linear acceleration seen by the accelerometer. Furthermore, this rotation is not consistent with the action of the human head following impact with the ground. The force plate in this st udy's setup measures the force acting on the skull only in the unprotected case. In the helmeted cases, the force plate measures the force acting on the outside of the helmet, not the skull itself. This does not provide useful information about the impac t force severity at the skull, which would be used to predict skull fracture.

PAGE 84

! "# CONCLUSIONS AND FUTURE RESEARCH According to the current study, it can be deduced that head injuries in the elderly population can be adequately protected a gainst using a combination of honeycomb and dilatent materials. Furthermore, the evidence presented here suggests that the prototype designed may be more effective than some current medical helmets. The impressive impact energy attenuation characteristic s of these materials are encouraging not only for the application of medical helmets, but also for bedside mats, recreational protection, etc. Further work in this field should include more comprehensive drop testing methods. In addition to the free fa ll drop test used here, the helmets should be tested while the headform is attached to the rest of the ATD and allowed to fall from bed sitting and standing scenarios. This will likely lead to increased repeatability issues and require more trials to gain sufficient confidence. Decreasing the number of drop heights, however, from the five used here to three will offset the added time associated with increasing the number of trials; keeping the study length manageable. The impact testing of material s phase could perhaps play a more prominent role in the design of the prototypes. For example, the impact energies associated with three fall severities could be quantified using the ATD. These three energies could be applied to the impact tower to deter mine the best materials for attenuating each impact energy level. Combining these materials in series in one prototype would create a three phase head protection device that would be well suited to many falling scenarios. This is presented as one of many approaches that should be considered. No matter the approach used, the impact testing of materials phase of research should play a more prominent role in the final prototype design. Future work might also quantify the areas under the force time curves p roduced during the impact tests. These areas are theoretically equal to "impulse" and should be equal across all materials. An ideal force time trace should be theorized and designed for

PAGE 85

! "# to minimize HIC15 while keeping peak force and acceleration under s ome acceptable limit. In this way, the design of future materials may be dictated/backed by solid engineering principles.

PAGE 86

! "# LITERATURE CITED 1 Rakier, A., Guilburd, J. N., Soustiel, J. F., Zaaroor, M., Feinsod, M. "Head Injuries in the Elderly." Br ain Injury 9 (1995): 187 194. 2 Tinnetti, M. E., Baker, D., McAvay, G., Claus, E. B., Garrett, P., Gottschalk, M., Koch, M. L., Trainor, K., Horwitz, R. I. "A Multifactorial Intervention to Reduce the Risk of Falling Among Elderly People Living in th e Community." New England Journal of Medicine 331 (1994): 821 827. 3 Nevitt, M. C. (1997). Ibid. 4 Thompson, H. J., McCormick, W. C., Kagan, S. H. "Traumatic Brain Injury in Older Adults: Epidemiology, Outcomes, and Future Implications." Journal o f the American Geriatric Society 54 (2006): 1590 1595. 5 Esquenazi, A. "Falls and Fractures in Older Post Stroke Patients with Spasticity: Consequences, and Drug Treatment Considerations." Clinical Geriatrics 12 (2004): 27 35. 6 Zoghi Moghadam, M ., Sadegh, A. M. "Global/local Head Models to Analyse Cerebral Blood Vessel Rupture Leading to ASDH and SAH." Computer Methods in Biomechanics and Biomedical Engineering 12 (2009): 1 12. 7 Hardman, J. M., Manoukian, A. "Pathology of Head Trauma." Ne uroimaging Clinics of North America 12 (2002): 175 187. 8 Medley, A., Thompson, M., French, J. "Predicting the Probability of Falls in Community Dwelling Pers ons With Brain Injury: A pilot study." Brain Injury 20 (2006): 1403 1408. 9 Tinetti, M. E. Speechley, M. "Prevention of Falls Among the Elderly." Medical Intelligence 320 (1989): 1055 1059. 10 Kim, K. J., Ashton Miller, J. A. "Biomechanics of Fall Arrest Using the Upper Extremity: Age differences." Clinical Biomechanics 18 (2003): 311 318. 11 Nigg, B. M., Herzog, W. "Biological Materials: Adaptation of Biological Materials to Excercise, Disuse, and Aging: Aging." Biomechanics of the Musculo Skeletal System England: Wiley, 2007. 232 235.

PAGE 87

! "# 12 Koskinen, S., Alaranta, H. "Traumat ic Brain Injury in Finland 1991 2005: A nationwide register study of hospitalized and fatal TBI." Brain Injury 22 (2008): 205 214. 13 Ownsworth, T., Little, T., Turner, B., Hawkes, A., Shum, D. "Assessing Emotional Status Following Acquired Brain In jury: The clinical potential of the depression, anxiety and stress scales." Brain Injury 22 (2008): 858 869. 14 Hendrich, A., Nyhuis, A., Kippenbrock, T., Soja, M. E. "Hospital Falls: Development of a predictive model for clinical practice." Applied Nursing Research 8 (1995): 129 139. 15 Yoganandan, N., Pintar, F. A. "Biomechanics of Temporo parietal Skull Fracture." Clinical Biomechanics 19 (2004): 225 239. 16 O'Riordain, K., et al. "Reconstruction of Real World Injury Accidents Resulting From Falls Using Mulitbody Dynamics." Clinical Biomechanics 18 (2003): 590 600. 17 Doorly, M.C., Gilchrist, M.D. "The Use of Accident Reconstruction for the Analysis of Traumatic Brain Injury Due to Head Impacts Arising From Falls." Computer Methods i n Biomechanics and Biomedical Engineering 9 (2006): 371 377. 18 McHenry, B. G. "Head Injury Criterion and the ATB." ATB Users' Group (2004): 1 8. 19 Yoganandan, N., Pintar, F. "Statistic ally and Biomechanically Based Criterion for Impact Induced Skull Fracture." 47th Annual Proceedings: Association for the Advancement of Automotive Medicine, 2003. 20 Eppinger, R., Sun, E., Kuppa, S., Saul, R. "Reports to National Highway Safety Administration." (2000). Supplement: Development of improved i njury criteria for the assessment of advanced automotive restraint systems II. Retrieved November 11, 2008 from http://www nrd.nhsta.dot.gov/pdf/nrd 11/airbags/finalrule_all.pdf 21 Chinn, B., Canaple, B., Derler, S. E., Doyle, D., Otte, D., Schuller, E., Willinger, R. "COST 327: Motorcycle Safety Helmets." Luxembourg: Office for Official Publications of the European Communities, 2001. 22 Ward, C., Chan, M., Nahum, A. "Intracranial Pressure: A brain injury criterion." Biomechanics of Impact In jury and Injury Tolerances of the Head Neck Complex Ed. Stanley H. Backaitis. Warrendale: Society of Automotive Engineers, 1993. 347 360.

PAGE 88

! "# 23 Nahum, A. M., Gatts, J. D., Gadd, C. W., Danforth, J. "Impact Tolerance of the Skull and Face." Biomechan ics of Impact Injury and Injury Tolerances of the Head Neck Complex Ed. Stanley H. Backaitis. Warrendale: Society of Automotive Engineers, 1993. 631 645. 24 Abele, J. R., Hyde, A. S., Cohen, H. H., LaRue, C. A., Bakken, G. M. "Stability, Fall an d Injury." Slips, Trips, Missteps and Their Consequences, Second Edition Tucson: Lawyers & Judges Publishing Company, Inc., 2005. 119 120. 25 Ford, D. "Mototcycle Helmet Performance: Blowing the Lid Off." Motorcyclist June 2005. 26 Thom, D. R "Comparison Tests of Motorcycle Helmets Qualified to International Standards." El Segundo: Collision and Injury Dynamics, 2006. 27 Withnall, C., Shewchenko, N., Wonnacott, M., Dvorak, J., Delaney, J. S. "Effectiveness of Headgear in Football." Bri tish Journal of Sports Medicine 39 (2005): i40 i48. 28 Shewchenko, N., Withnall, C., Keown, M., Gittens, R., Dvorak, J. "Heading in Football. Part 1: Development of biomechanical methods to investigate head response." British Journal of Sports Medici ne 39 (2005): i10 i25. 29 Shewchenko, N., Withnall, C., Keown, M., Gittens, R., Dvorak, J. "Heading in Football. Part 2: Biomechanics of ball heading and head response." British Journal of Sports Medicine 39 (2005): i26 i32. 30 Withnall, C., Shewch enko, N., Gittens, R., Dvorak, J. "Biomechanical Investigation of Head Impacts in Football." British Journal of Sports Medicine 39 (2005): i49 i57. 31 Sherker, S., Ozanne Smith, J. "Are Current Playground Safety Adequate for Preventing Arm Fracture s?" MJA 180 (2004): 562 565. 32 Gunatilaka, A. H., Sherker, S., Ozanne Smith, J. "Comparative Performance of Playground Surfacing Materials Including Conditions of Extreme Non Compliance." Injury Prevention 10 (2004): 174 179. 33 Knouse, C. L., G ould, T. E., Caswell, S. V., Deivert, R. G. "Efficacy of Rugby Headgear in Attenuating Repetitive Linear Impact Forces." Journal of Athletic Training 38 (2003): 330 335. 34 Mills, N. J. "Protective Capability of Bicycle Helmets." British Journal of Sports Medicine 24 (1990): 55 60.

PAGE 89

! "# APPENDIX 1: MATLAB C ODE

PAGE 90

! "# Appendix 1 % Impact testing of i ndividual Material s -----------------clc clear % Load all raw data load A.txt load D.txt load D1.txt load DCR.t xt load DCW.txt load DCY.txt load E.txt load F25.txt load F25E.txt load F33.txt load G.txt load H.txt load S2.txt load Unprotected.txt % Define all variables Atime = A(:,1); Aa = A(:,2); Ab = A(:,3); Ac = A(:,4); Ad = A(:,5); Ae = A(:,6); Dtime = D(:,1 ); Da = D(:,2); Db = D(:,3); Dc = D(:,4); Dd = D(:,5); De = D(:,6); D1time = D1(:,1); D1a = D1(:,2); D1b = D1(:,3); D1c = D1(:,4); D1d = D1(:,5); D1e = D1(:,6); DCRtime = DCR(:,1); DCRa = DCR(:,2); DCRb = DCR(:,3);

PAGE 91

! "# Appendix 1 (Continued) DCRc = DCR(:, 4); DCRd = DCR(:,5); DCRe = DCR(:,6); DCWtime = DCW(:,1); DCWa = DCW(:,2); DCWb = DCW(:,3); DCWc = DCW(:,4); DCWd = DCW(:,5); DCWe = DCW(:,6); DCYtime = DCY(:,1); DCYa = DCY(:,2); DCYb = DCY(:,3); DCYc = DCY(:,4); DCYd = DCY(:,5); DCYe = DCY(:,6); Et ime = E(:,1); Ea = E(:,2); Eb = E(:,3); Ec = E(:,4); Ed = E(:,5); Ee = E(:,6); F25time = F25(:,1); F25a = F25(:,2); F25b = F25(:,3); F25c = F25(:,4); F25d = F25(:,5); F25e = F25(:,6); F25Etime = F25E(:,1); F25Ea = F25E(:,2); F25Eb = F25E(:,3); F25Ec = F25E(:,4); F25Ed = F25E(:,5); F25Ee = F25E(:,6); F33time = F33(:,1); F33a = F33(:,2); F33b = F33(:,3); F33c = F33(:,4); F33d = F33(:,5); F33e = F33(:,6);

PAGE 92

! "" Appendix 1 (Continued) Gtime = G(:,1); Ga = G(:,2); Gb = G(:,3); Gc = G(:,4); Gd = G(:,5); Ge = G( :,6); Htime = H(:,1); Ha = H(:,2); Hb = H(:,3); Hc = H(:,4); Hd = H(:,5); He = H(:,6); S2time = S2(:,1); S2a = S2(:,2); S2b = S2(:,3); S2c = S2(:,4); S2d = S2(:,5); S2e = S2(:,6); Unprotectedtime = Unprotected(:,1); Unprotected = Unprotected(:,2); % Filter all force sensor data (1800 Hz cutoff as was used for force plate sensors) n = 8192; % Number of samples taken per test SamplingDuration = 0.05; % Time duration (sec) of sampling Fs = n/SamplingDuration ; % Sampling frequency in Hz(Fs) N = 4; % Filter order (N) Fc = 1800; % Cutoff frequency [B,A] = butter(N,Fc/(Fs/2)); % Find transfer function coefficients from Butterw orth filter parameters Aa_filt = filtfilt(B,A,Aa); % Forward backward filtering of raw sensor data according to butterworth transfer function Ab_filt = filtfilt(B,A,Ab); Ac_filt = filtfilt(B,A,Ac); Ad_filt = filtfilt(B,A,Ad); Ae_filt = filtfilt(B ,A,Ae);

PAGE 93

! "# Appendix 1 (Continued) vgA_filt = (Aa_filt+Ab_filt+Ac_filt+Ad_filt+Ae_filt)./5; % Take average of five trials max(avgA_filt); % Find peak force Da_filt = filtfilt(B,A,Da); Db_filt = filtfilt(B,A,Db); Dc_filt = filtfilt(B,A,Dc); Dd_filt = filtfilt(B,A,Dd); De_filt = filtfilt(B,A,De); avgD_filt = (Da_filt+Db_filt+Dc_filt+Dd_filt+De_filt)./5; max(avgD_filt); D1a_filt = filtfilt(B,A,D1a); D1b_filt = filtfilt(B,A,D1b); D1c_filt = filtfilt(B,A,D1c); D1d_ filt = filtfilt(B,A,D1d); D1e_filt = filtfilt(B,A,D1e); avgD1_filt = (D1a_filt+D1b_filt+D1c_filt+D1d_filt+D1e_filt)./5; max(avgD1_filt); DCRa_filt = filtfilt(B,A,DCRa); DCRb_filt = filtfilt(B,A,DCRb); DCRc_filt = filtfilt(B,A,DCRc); DCRd_filt = filtfilt( B,A,DCRd); DCRe_filt = filtfilt(B,A,DCRe); avgDCR_filt = (DCRa_filt+DCRb_filt+DCRc_filt+DCRd_filt+DCRe_filt)./5; max(avgDCR_filt); DCWa_filt = filtfilt(B,A,DCWa); DCWb_filt = filtfilt(B,A,DCWb); DCWc_filt = filtfilt(B,A,DCWc); DCWd_filt = filtfilt(B,A,DC Wd); DCWe_filt = filtfilt(B,A,DCWe); avgDCW_filt = (DCWa_filt+DCWb_filt+DCWc_filt+DCWd_filt+DCWe_filt)./5; max(avgDCW_filt); DCYa_filt = filtfilt(B,A,DCYa); DCYb_filt = filtfilt(B,A,DCYb); DCYc_filt = filtfilt(B,A,DCYc); DCYd_filt = filtfilt(B,A,DCYd); D CYe_filt = filtfilt(B,A,DCYe);

PAGE 94

! "# Appendix 1 (Continued) avgDCY_filt = (DCYa_filt+DCYb_filt+DCYc_filt+DCYd_filt+DCYe_filt)./5; max(avgDCY_filt); Ea_filt = filtfilt(B,A,Ea); Eb_filt = filtfilt(B,A,Eb); Ec_filt = filtfilt(B,A,Ec); Ed_filt = filtfilt(B,A,Ed); Ee_filt = filtfilt(B,A,Ee); avgE_filt = (Ea_filt+Eb_filt+Ec_filt+Ed_filt+Ee_filt)./5; max(avgE_filt); F25a_filt = filtfilt(B,A,F25a); F25b_filt = filtfilt(B,A,F25b); F25c_filt = filtfilt(B,A,F25c); F25d_filt = filtfilt(B,A,F25d); F25e_filt = filtfilt(B, A,F25e); avgF25_filt = (F25a_filt+F25b_filt+F25c_filt+F25d_filt+F25e_filt)./5; max(avgF25_filt); F25Ea_filt = filtfilt(B,A,F25Ea); F25Eb_filt = filtfilt(B,A,F25Eb); F25Ec_filt = filtfilt(B,A,F25Ec); F25Ed_filt = filtfilt(B,A,F25Ed); F25Ee_filt = filtfilt (B,A,F25Ee); avgF25E_filt = (F25Ea_filt+F25Eb_filt+F25Ec_filt+F25Ed_filt+F25Ee_filt)./5 ; max(avgF25E_filt); F33a_filt = filtfilt(B,A,F33a); F33b_filt = filtfilt(B,A,F33b); F33c_filt = filtfilt(B,A,F33c); F33d_filt = filtfilt(B,A,F33d); F33e_filt = filtfi lt(B,A,F33e); avgF33_filt = (F33a_filt+F33b_filt+F33c_filt+F33d_filt+F33e_filt)./5; max(avgF33_filt); Ga_filt = filtfilt(B,A,Ga); Gb_filt = filtfilt(B,A,Gb); Gc_filt = filtfilt(B,A,Gc); Gd_filt = filtfilt(B,A,Gd); Ge_filt = filtfilt(B,A,Ge);

PAGE 95

! "# Appendix 1 (Continued) avgG_filt = (Ga_filt+Gb_filt+Gc_filt+Gd_filt+Ge_filt)./5; max(avgG_filt); _filt = filtfilt(B,A,Ha); Hb_filt = filtfilt(B,A,Hb); Hc_filt = filtfilt(B,A,Hc); Hd_filt = filtfilt(B,A,Hd); He_filt = filtfilt(B,A,He); avgH_filt = (Ha_filt+Hb_filt+ Hc_filt+Hd_filt+He_filt)./5; max(avgH_filt); S2a_filt = filtfilt(B,A,S2a); S2b_filt = filtfilt(B,A,S2b); S2c_filt = filtfilt(B,A,S2c); S2d_filt = filtfilt(B,A,S2d); S2e_filt = filtfilt(B,A,S2e); avgS2_filt = (S2a_filt+S2b_filt+S2c_filt+S2d_filt+S2e_filt) ./5; max(avgS2_filt); Unprotected_filt = filtfilt(B,A,Unprotected); max(Unprotected_filt); plot(Unprotectedtime,Unprotected_filt,Atime,avgA_filt,Dtime ,avgD_filt,D1time,avgD1_filt,DCRtime,avgDCR_filt,DCWtime,av gDCW_filt,DCYtime,avgDCY_filt,Etime,avgE_fi lt,F25time,avgF2 5_filt,F25Etime,avgF25E_filt,F33time,avgF33_filt,Gtime,avgG _filt,Htime,avgH_filt,S2time,avgS2_filt); % Convert force data to acceleration (G's) using F=ma and normalizing by G m = 6.3503; % Drop weight mass(kg) G = 9.81; % Accleration due to gravity (m/s^2) aAa = (Aa_filt./m)./G; aAb = (Ab_filt./m)./G; aAc = (Ac_filt./m)./G; aAd = (Ad_filt./m)./G; aAe = (Ae_filt./m)./G; aavgA = (aAa+aAb+aAc+aAd+aAe)./5; % Take average of 5 tests max(aavgA); % plot(Atim e,aavgA)

PAGE 96

! "# Appendix 1 (Continued) aDa = (Da_filt./m)./G; aDb = (Db_filt./m)./G; aDc = (Dc_filt./m)./G; aDd = (Dd_filt./m)./G; aDe = (De_filt./m)./G; aavgD = (aDa+aDb+aDc+aDd+aDe)./5; max(aavgD); % plot(Dtime,aavgD) aD1a = (D1a_filt./m)./G; aD1b = (D1b_filt./m)./G; aD1c = (D1c_filt./m)./G; aD1d = (D1d_filt./m)./G; aD1e = (D1e_filt./m)./G; aavgD1 = (aD1a+aD1b+aD1c+aD1d+aD1e)./5; max(aavgD1); % plot(D1time,aavgD1) aDCRa = (DCRa_filt./m)./G; aDCRb = (DCRb_filt./m)./G; aDCRc = (DCRc_filt./m)./G; aD CRd = (DCRd_filt./m)./G; aDCRe = (DCRe_filt./m)./G; aavgDCR = (aDCRa+aDCRb+aDCRc+aDCRd+aDCRe)./5; max(aavgDCR); % plot(DCRtime,aavgDCR) aDCWa = (DCWa_filt./m)./G; aDCWb = (DCWb_filt./m)./G; aDCWc = (DCWc_filt./m)./G; aDCWd = (DCWd_filt./m)./G; aDCWe = (DCWe_filt./m)./G; aavgDCW = (aDCWa+aDCWb+aDCWc+aDCWd+aDCWe)./5; max(aavgDCW); % plot(DCWtime,aavgDCW) aDCYa = (DCYa_filt./m)./G; aDCYb = (DCYb_filt./m)./G; aDCYc = (DCYc_filt./m)./G; aDCYd = (DCYd_filt./m)./G; aDCYe = (DCYe_filt./m)./G;

PAGE 97

! "# Appendix 1 (Cont inued) aavgDCY = (aDCYa+aDCYb+aDCYc+aDCYd+aDCYe)./5; max(aavgDCY); % plot(DCRtime,aavgDCR) aEa = (Ea_filt./m)./G; aEb = (Eb_filt./m)./G; aEc = (Ec_filt./m)./G; aEd = (Ed_filt./m)./G; aEe = (Ee_filt./m)./G; aavgE = (aEa+aEb+aEc+aEd+aEe)./5; max(aavgE); %plot(Etime,aavgE); aF25a = (F25a_filt./m)./G; aF25b = (F25b_filt./m)./G; aF25c = (F25c_filt./m)./G; aF25d = (F25d_filt./m)./G; aF25e = (F25e_filt./m)./G; aavgF25 = (aF25a+aF25b+aF25c+aF25d+aF25e)./5; max(aavgF25); %plot(F25time,aavgF25); aF25Ea = (F25Ea_filt./m)./G; aF25Eb = (F25Eb_filt./m)./G; aF25Ec = (F25Ec_filt./m)./G; aF25Ed = (F25Ed_filt./m)./G; aF25Ee = (F25Ee_filt./m)./G; aavgF25E = (aF25Ea+aF25Eb+aF25Ec+aF25Ed+aF25Ee)./5; max(aavgF25E); %plot(F25Etime,aavgF25E); aF33a = (F33a_filt./m). /G; aF33b = (F33b_filt./m)./G; aF33c = (F33c_filt./m)./G; aF33d = (F33d_filt./m)./G; aF33e = (F33e_filt./m)./G; aavgF33 = (aF33a+aF33b+aF33c+aF33d+aF33e)./5; max(aavgF33); %plot(D1F33time,aavgD1F33);

PAGE 98

! "# Appendix 1 (Continued) aGa = (Ga_filt./m)./G; aGb = (Gb_filt./m)./G; aGc = (Gc_filt./m)./G; aGd = (Gd_filt./m)./G; aGe = (Ge_filt./m)./G; aavgG = (aGa+aGb+aGc+aGd+aGe)./5; max(aavgG); %plot(Gtime,aavgG); aHa = (Ha_filt./m)./G; aHb = (Hb_filt./m)./G; aHc = (Hc_filt./m)./G; aHd = (Hd_filt./m)./G; aHe = ( He_filt./m)./G; aavgH = (aHa+aHb+aHc+aHd+aHe)./5; max(aavgH); %plot(D1Htime,aavgD1H); aS2a = (S2a_filt./m)./G; aS2b = (S2b_filt./m)./G; aS2c = (S2c_filt./m)./G; aS2d = (S2d_filt./m)./G; aS2e = (S2e_filt./m)./G; aavgS2 = (aS2a+aS2b+aS2c+aS2d+aS2e)./5; m ax(aavgS2); %plot(S2time,aavgS2); aUnprotected = (Unprotected_filt./m)./G; max(aUnprotected); % plot(Unprotectedtime,aUnprotected) % plot(Atime,aavgA,Dtime,aavgD,D1time,aavgD1,DCRtime,aavgDCR, DCWtime,aavgDCW,DCYtime,aavgDCY,Etime,aavgE,F25time,aavgF2 5 ,F25Etime,aavgF25E,F33time,aavgF33,Gtime,aavgG,Htime,aavgH, S2time,aavgS2,Unprotectedtime,aUnprotected) % HIC Calculation (15 ms) % dt = 164:164:2460; % for j=1:length(dt)

PAGE 99

! "# Appendix 1 (Continued) % for i=1:5732 % Ahic15(i,j) = (((1/(dt(j)/F s))*trapz(Atime(i:(i+dt(j))),aavgA(i:(i+dt(j)) )))^2.5)*(dt(j)/Fs); % end % end % AHIC15 = max(Ahic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % Dhic15(i,j) = (((1/(dt(j)/Fs))*trapz(Dtime(i:(i+dt(j))),aavgD(i:(i+dt(j)) )))^2.5) *(dt(j)/Fs); % end % end % DHIC15 = max(Dhic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % D1hic15(i,j) = (((1/(dt(j)/Fs))*trapz(D1time(i:(i+dt(j))),aavgD1(i:(i+dt(j )))))^2.5)*(dt(j)/Fs); % end % end % D1HIC15 = max(D1hic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % DCRhic15(i,j) = (((1/(dt(j)/Fs))*trapz(DCRtime(i:(i+dt(j))),aavgDCR(i:(i+dt (j)))))^2.5)*(dt(j)/Fs); % end % end % DCRHIC15 = max(DCRhic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % DCWhic15(i,j) = (((1/(dt(j)/Fs))*trapz(DCWtime(i:(i+dt(j))),aavgDCW(i:(i+dt (j)))))^2.5)*(dt(j)/Fs);

PAGE 100

! "# Appendix 1 (Continued) % end % end % DCWHIC15 = max(DCWhic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % DCYhic15(i,j) = (((1/(dt(j)/Fs))*trapz(DCYtime(i:(i+dt(j))),aavgDCY(i:(i+dt (j)))))^2.5)*(dt(j)/Fs); % end % end % DCYHIC15 = max(DCYhic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % Ehic15(i,j) = (((1/(dt(j)/Fs))*trapz(Etime(i: (i+dt(j))),aavgE(i:(i+dt(j)) )))^2.5)*(dt(j)/Fs); % end % end % EHIC15 = max(Ehic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % F25hic15(i,j) = (((1/(dt(j)/Fs))*trapz(F25time(i:(i+dt(j))),aavgF25(i:(i+dt (j)))))^2.5)*(dt(j)/Fs); % end % end % F25HIC15 = max(F25hic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % F25Ehic15(i,j) = (((1/(dt(j)/Fs))*trapz(F25Etime(i:(i+dt(j))),aavgF25E(i:(i+ dt(j)))))^2.5)*(dt(j)/Fs); % end % end % F25EHIC15 = max(F25Ehic15) % dt = 164:164:2460;

PAGE 101

! "# Appendix 1 (Continued) % for j=1:length(dt) % for i=1:5732 % F33hic15(i,j) = (((1/(dt(j)/Fs))*trapz(F33time(i:(i+dt(j))),aavgF33(i:(i+dt (j)))))^2.5)*(dt(j)/Fs); % end % end % F33HIC15 = max(F33hic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % Ghic15(i,j) = (((1/(dt(j)/Fs))*trapz(Gtime(i:(i+dt(j))),aavgG(i:(i+dt(j)) )))^2.5)*(dt(j)/Fs); % end % end % GHIC15 = max(Ghic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % Hhic15(i, j) = (((1/(dt(j)/Fs))*trapz(Htime(i:(i+dt(j))),aavgH(i:(i+dt(j)) )))^2.5)*(dt(j)/Fs); % end % end % HHIC15 = max(Hhic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % S2hic15(i,j) = (((1/(dt(j)/Fs))*trapz(S2time(i:(i+dt(j))),aavgS2( i:(i+dt(j )))))^2.5)*(dt(j)/Fs); % end % end % S2HIC15 = max(S2hic15) % dt = 164:164:2460; % for j=1:length(dt) % for i=1:5732 % Unprotectedhic15(i,j) = (((1/(dt(j)/Fs))*trapz(Unprotectedtime(i:(i+dt(j))),aUnprot ected(i:(i+dt(j)))))^2.5)*(dt(j)/ Fs);

PAGE 102

! "# Appendix 1 (Continued) % end % end % UnprotectedHIC15 = max(Unprotectedhic15) % Drop testing -------------------------------------------% Rename data file of trial of interest "temp.txt" clc clear clear all warning( 'off' 'all' ) load te mp.txt % Define all variables sample = [1:1:24000]; headAP = temp(:,1); headML = temp(:,2); headSI = temp(:,3); F_A = (temp(:,4))./.1464; F_B = (temp(:,5))./.15; F_C = (temp(:,6))./.1456; F_D = (temp(:,7))./.1414; % Filter headform data (1650 Hz is f or headform data Find appropriate cutoff frequency for FP data) Fs = 12000; % Sampling frequency (Fs) N = 4; % Filter order number (N) Fc_head = 1650; Fc_FP = 1800; % Cutoff frequenci es [B,A] = butter(N,Fc_head/(Fs/2)); % Find transfer function coefficients from Butterworth filter parameters [D,C] = butter(N,Fc_FP/(Fs/2)); % Find transfer function coefficients from Butterworth filter parameters headAP_filt = filtfilt(B, A,headAP); % Forward backward filtering of raw headform data according to butterworth transfer function headML_filt = filtfilt(B,A,headML); headSI_filt = filtfilt(B,A,headSI);

PAGE 103

! "" Appendix 1 (Continued) % % Remove offset Find avg of first 10,000 samples and subtract from signals % cfAP = sum(headAP_filt(1:10000))/10000; % cfML = sum(headML_filt(1:10000))/10000; % cfSI = sum(headSI_filt(1:10000))/10000; % % headAP_corrected = headAP_filt cfAP; % headML_corrected = headML_filt cfML; % headSI_corrected = h eadSI_filt cfSI; F_A_filt = filtfilt(D,C,F_A); % Forward backward filtering of raw FP data according to butterworth transfer function F_B_filt = filtfilt(D,C,F_B); F_C_filt = filtfilt(D,C,F_C); F_D_filt = filtfilt(D,C,F_D); for i=1:length(sam ple) head_resultant_filt(i) = ((headAP_filt(i)^2)+(headML_filt(i)^2)+(headSI_filt(i)^2))^ .5; FP_resultant_filt(i) = F_A_filt(i)+F_B_filt(i)+F_C_filt(i)+F_D_filt(i); end % %remove secondary impacts from data head_resultant_filt(15000:24000)=0; FP_ resultant_filt(15000:24000)=0; % F_A_filt(13000:24000)=0; % F_B_filt(13000:24000)=0; % F_C_filt(13000:24000)=0; % F_D_filt(13000:24000)=0; % plot(sample,headAP_correct,sample,headML_correct,sample,hea dSI_correct) % plot(sample,F_A_filt,sample,F_B_filt,sa mple,F_C_filt,sample ,F_D_filt) plot(sample,head_resultant_filt) % plot(sample,FP_resultant_filt) HeadMAX = max(head_resultant_filt) FPMAX = max(FP_resultant_filt) % FAmax = max(F_A_filt)

PAGE 104

! "# Appendix 1 (Continued) % FBmax = max(F_B_filt) % FCmax = max(F_C_f ilt) % FDmax = max(F_D_filt) % % Comparison of Raw and Filtered data % plot(sample,headAP,sample,headAP_filt) % title('Comparison of Raw and Filtered Data') % xlabel('Sample #') % ylabel('G') % display 'Peak G:' % max(abs(headAP_filt)) % % Amplitude S pectrum Comparison of the four force plate load cells % Fs = 12000; % Sampling frequency is 12,000 (Hz) % T = 1/Fs; % Time between each sample taken (sec) % L = length(sample); % Length of signal % t = (0:L 1)*T; % Time vector (sec) % NFFT = 2^nextpow2(L); % Next power of 2 from length of y % f = Fs/2*linspace(0,1,NFFT/2); % Frequency vector created % % A = fft(F_A,NFFT)/L; % FFT of raw signal % B = fft(F_B,N FFT)/L; % C = fft(F_C,NFFT)/L; % D = fft(F_D,NFFT)/L; % % A_filt = fft(F_A_filt,NFFT)/L; % FFT of filtered signal % B_filt = fft(F_B_filt,NFFT)/L; % C_filt = fft(F_C_filt,NFFT)/L; % D_filt = fft(F_D_filt,NFFT)/L; % % subplot(4,1,1); plot(f,2*abs(A (1:NFFT/2)),f,2*abs(A_filt(1:NFFT/2))) % Plot single sided amplitude spectrum % title('Single Sided Amplitude Spectrum') % xlabel('Frequency (Hz)') % ylabel('|A(f)|') % subplot(4,1,2); plot(f,2*abs(B(1:NFFT/2)),f,2*abs(B_filt(1:NFFT/2))) % ylabel('|B( f)|')

PAGE 105

! "# Appendix 1 (Continued) % subplot(4,1,3); plot(f,2*abs(C(1:NFFT/2)),f,2*abs(C_filt(1:NFFT/2))) % ylabel('|C(f)|') % subplot(4,1,4); plot(f,2*abs(D(1:NFFT/2)),f,2*abs(D_filt(1:NFFT/2))) % ylabel('|D(f)|') % % Visualize head filter's magnitude and p hase behavior % Wn = 1650/(12000/2); % Wn in butter function is cutoff freq divided by half of the sampling freq % [z,p,k] = butter(4,Wn); % Gives zeros, poles, and gain for the butterworth filter corresponding to the input info % [sos,g ] = zp2sos(z,p,k); % Convert to SOS form % Hd = dfilt.df2tsos(sos,g); % Create a dfilt object % h = fvtool(Hd); % Plot magnitude response % set(h,'Analysis','freq') % Display frequency response % % Visualize force plat e filter's magnitude and phase behavior % Wn = 1800/(12000/2); % Wn in butter function is cutoff freq divided by half of the sampling freq % [z,p,k] = butter(4,Wn); % Gives zeros, poles, and gain for the butterworth filter corresponding to the input info % [sos,g] = zp2sos(z,p,k); % Convert to SOS form % Hd = dfilt.df2tsos(sos,g); % Create a dfilt object % h = fvtool(Hd); % Plot magnitude response % set(h,'Analysis','freq') % Display frequency response % HIC (15 ms) dt = 0.001:0.001:0.015; start = 200; finish = 24000; for j=1:length(dt) for i=start:finish hic15(i,j) = (((1/dt(j))*trapz(sample((i dt(j)*Fs):i)/Fs,head_resultant_filt(((i dt(j)*Fs):i))))^2.5)*dt(j);

PAGE 106

! "# Appendix 1 (Continued) end end HIC15 = max(max(hic15)) % % HIC (36 ms) % dt = 0.001:0.001:0.036; % start = 19000; % finish = 24000; % for j=1:length(dt) % for i=start:finish % hic36(i,j) = (((1/dt(j))*trapz(sample((i dt(j)*Fs):i)/Fs,head_resultant_filt(((i dt(j)* Fs):i))))^2.5)*dt(j); % end % end % % HIC36 = max(hic36)

PAGE 107

! "# APPENDIX 2: SUPPLEME NTARY RESULTS

PAGE 108

! "# Appendix 2 Figure 21 : Comparison of 3 prototype designs (peak acceleration in the back orientation) Figure 22 : Comparison of 3 prototype designs (peak acceleration in the side orientation)

PAGE 109

! "# Appendix 2 (Continued) Figure 23 : Comparison of 3 prototype designs (peak force i n the back orientation) Figure 24 : Comparison of 3 prototype designs (peak force in the side orientation)

PAGE 110

! "# Appendix 2 (Continued) Figure 25 : Comparison of 3 prototype designs (HIC15 in the back o rientation) Figure 26 : Comparison of 3 prototype designs (HIC15 in the side orientation)

PAGE 111

! "# Appendix 2 (Continued) Figure 27 : Comparison of "Prototype 1" and 2 medical helmets (peak acceleration in the back orientation) Figure 28 : Comparison of "Prototype 1" and 2 medical hel mets (peak acceleration in the side orientation)

PAGE 112

! "# Appendix 2 (Continued) Figure 29 : Comparison of "Prototype 1" and 2 medical helmets (peak force in the back orientation) Figure 30 : Comparison of "Prototype 1" and 2 medical helmets (peak force in the side orientation)

PAGE 113

! "# Appendix 2 (Continued) Figure 31 : Comparison of "Prototype 1" and 2 medical helmets (HIC15 in the back orientation) Figure 32 : Comparison of "Prototype 1" and 2 medical helmets (HIC15 in the side orientation)

PAGE 114

! "" Appendix 2 (Continued) Figure 33 : Comparison of "Prototype 1" to the unprotected case (HIC15 in the back orientation) Figure 34 : Comparison of "Prototype 1" to the unprotected case (HIC15 in the side orientation)

PAGE 115

! "## Appendix 2 (Continued) Figure 35 : Comparison of "Prototype 1" to the unprotected case (peak acceleration in the back orientation) ! Figure 36 : Comparison of "Prototype 1" to the unprotected case (peak acceleration in the side orientation)

PAGE 116

! "#" App endix 2 (Continued) Table 13 : Impact testing results of individual materials

PAGE 117

! "#$ Appendix 2 (Continued) Table 14 : Impact testing results of material combinations

PAGE 118

! "#$ Appen dix 2 (Continued) Table 15 : Head drop testing results of the unprotected case

PAGE 119

! "#$ Appendix 2 (Continued) Table 15 (Continued)

PAGE 120

! "#$ Appendix 2 (Continued) Table 16 : Head drop testing results for "Prototype 1"

PAGE 121

! "#$ Appendix 2 (Continued) Table 17 : Head drop testing results for "Prototype 2"

PAGE 122

! "#$ Appendix 2 (Continued) Table 18 : Prototype 3 head drop testing results

PAGE 123

! "#$ Appendix 2 (Continued ) Table 19 : Head drop testing results for Plum ProtectaCap+Plus medical helmet

PAGE 124

! "#$ Appendix 2 (Continued) Table 20 : Head drop testing re sults for HipSaver medical helmet

PAGE 125

! ""# APPENDIX 3: PHOTOGRA PHS OF MATERIALS & M EDICAL HELMETS

PAGE 126

! """ Appendix 3 Figure 37 : "DC W" dilatent material (front view) ! Figure 38 : "DC W" dilatent material (back view)

PAGE 127

! ""# Appendix 3 (Continued) ! Figure 39 : "A" dilatent material (front view) Figure 40 : "A" dilatent material (back view)

PAGE 128

! ""# Appendix 3 (Continued) ! Figure 41 : "S2" dilatent material, also referred to as "Dd" (front view) Figure 42 : "S2" dilatent material, also referred to as "Dd" (back view)

PAGE 129

! ""# Appendix 3 (Continued) Figure 43 : "B" dilatent material (front view) Figure 44 : "B" dilatent material (back view)

PAGE 130

! ""# Appendix 3 (Continued) Figure 45 : "C" dilatent material (front view) Figure 46 : "C" dilatent material (back view)

PAGE 131

! ""# Appendix 3 (Continued) Figure 47 : "DC R" dilatent material (front view) Figure 48 : "DC R" dilatent material (back view)

PAGE 132

! ""# Appendix 3 (Continued) Figure 49 : "S7" dilatent material (front view) Figure 50 : "S7" dilatent material (back view)

PAGE 133

! ""# Appendix 3 (Continued) Figure 51 : "DC Y" dilatent material (front view) ! Figure 52 : "DC Y" dilatent material (back view)

PAGE 134

! ""# Appendix 3 (Continued) Figure 53 : "D1" dilatent material (front view) Figure 54 : "D1" dilatent material (back view)

PAGE 135

! "#$ Appendix 3 (Continued) Figure 55 : "F25" honeycomb material, also referred to as "Hz" (front view) Figure 56 : "F25" honeycomb material, also referred to as "Hz" (back view)

PAGE 136

! "#" Appendix 3 (Continued) Figure 57 : "K" honeycomb material (front view) Figure 58 : "K" honeycomb material (back view)

PAGE 137

! "## Appendix 3 (Continued) Figure 59 : "F33" honeycomb material (front view) Figure 60 : "F33" honeycomb material (back view)

PAGE 138

! "#$ Appendix 3 (Continued) Figure 61 : "D" honeycomb material, also referred to as "Hy" (front view) ! Figure 62 : "D" honeycomb material, also referred to as "Hy" (back view) !

PAGE 139

! "#$ Appendix 3 (Continued) Figure 63 : "L" honeycomb material (front view) Figure 64 : "L" honeycomb material (back view)

PAGE 140

! "#$ Appendix 3 (Continued) Figure 65 : "J" honeycomb material (front view) Figure 66 : "J" honeycomb material (back view)

PAGE 141

! "#$ Appendix 3 (Continued) Figure 67 : "H" honeycomb material (front view) Figure 68 : "H" honeycomb material (back view)

PAGE 142

! "#$ Appendix 3 (Continued) Figure 69 : "E" honeycomb material (front view) Figure 70 : "E" honeycomb material (back view)

PAGE 143

! "#$ Append ix 3 (Continued) Figure 71 : "I" honeycomb material (front view) Figure 72 : "I" honeycomb material (back view)

PAGE 144

! "#$ Appendix 3 (Continued) Figure 73 : "G" honey comb material (front view) Figure 74 : "G" honeycomb material (back view)

PAGE 145

! "#$ Appendix 3 (Continued) Figure 75 : "F25E" honeycomb material (front view) Figure 76 : "F25E" honeyc omb material (back view)

PAGE 146

! "#" Appendix 3 (Continued) Figure 77 : "D2" dilatent material (front view) Figure 78 : "D2" dilatent material (back view)

PAGE 147

! "#$ Appendix 3 (Continued) Figure 79 : HipSaver medical helmet !

PAGE 148

! "## Appendix 3 (Continued) Figure 80 : Plum ProtectaCap+Plus medical helmet


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200397Ka 4500
controlfield tag 001 002069237
005 20100420104519.0
007 cr mnu|||uuuuu
008 100420s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003103
035
(OCoLC)608021951
040
FHM
c FHM
049
FHMM
090
TP145 (Online)
1 100
Kerrigan, Michael V.
0 245
Evaluation of advanced materials to protect against fall-related head injuries
h [electronic resource] /
by Michael V. Kerrigan.
260
[Tampa, Fla] :
b University of South Florida,
2009.
500
Title from PDF of title page.
Document formatted into pages; contains 133 pages.
502
Thesis (M.S.B.E.)--University of South Florida, 2009.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
3 520
ABSTRACT: Falls among the elderly population continue to be a growing concern in the healthcare industry and are marked by staggeringly high social and economic costs. The incidence of falls is known to increase with age, and currently the elderly population is growing at an astounding rate as baby-boomers are now entering this age group. Also, recovery following fall-related injuries decreases with increased age. These confounding factors currently make falls a very important area of research. Of the injuries typically seen in falls among the elderly, head injuries are one of the most debilitating. Death due to head trauma among the elderly is gaining national attention; head trauma is now considered the number one cause of death among elders who fall1. Among other technologies, medical helmets are often employed to protect against such injuries, but patient compliance with these helmets remains an issue. Current helmets use foams and cotton as padding, contributing to clumsy designs. Dilatent and honeycomb materials may be the future of this industry as their low weight and high efficacy per thickness make them ideal materials for thinner, lighter, less cumbersome head protection devices. This study outlines various modes of head injury and then highlights several head protection measures. The newer materials are tested using various methods to determine the most promising candidates for prototype designs. Next, three prototypes are assembled from the newer materials and compared directly based on the protection measures established. Finally, the top-performing prototype is compared against two existing medical helmets in a similar fashion. The results show that the best prototype significantly outperforms one of the existing medical helmets, and shows slight improvement over the other. These results establish the promise of these newer materials in the application of head protection devices.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Co-advisor: John D. Lloyd, Ph.D.
Co-advisor: William E. Lee III, Ph.D.
653
Falls
Biomechanics
Accelerometer
Head Injury Criteria (HIC)
Traumatic Brain Injury (TBI)
690
Dissertations, Academic
z USF
x Chemical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3103