A Thesis by. Oriol Oliva-Perez. Bachelor s of Science, Polytechnic University of Catalonia, 2000

Transcription

1 EVALUATION OF THE FAA HYBRID III 50 TH PERCENTILE ANTHROPOMETRIC TEST DUMMY UNDER THE FAR AND EMERGENCY LANDING CONDITIONS FOR THE COMBINED HORIZONTAL-VERTICAL DYNAMIC LOADING A Thesis by Oriol Oliva-Perez Bachelor s of Science, Polytechnic University of Catalonia, 2000 Submitted to the Department of Mechanical Engineering and the faculty of the Graduate School of Wichita State University in partial fulfillment of the requirements for the degree of Master of Science May 2010

2 Copyright 2010 by Oriol Oliva-Perez All Rights Reserved Note that thesis and dissertation work is protected by copyright, with all rights reserved. Only the author has the legal right to publish, produce, sell, or distribute this work. Author permission is needed for others to directly quote significant amounts of information in their own work to summarize substantial amounts of information in their own work. Limited amounts of information cited, paraphrased, or summarized from the work may be used with proper citation of where to find the original work.

3 EVALUATION OF THE FAA HYBRID III 50 TH PERCENTILE ANTHROPOMETRIC TEST DUMMY UNDER THE FAR AND EMERGENCY LANDING CONDITIONS FOR THE COMBINED HORIZONTAL-VERTICAL DYNAMIC LOADING The following faculty members have examined the final copy of this thesis for form and content, and recommend that it be accepted in partial fulfilment of the requirement for the degree of Master of Science with a major in Mechanical Engineering Hamid Lankarani, Committee Chair George Talia, Committee Co-Chair Gerardo Olivares, Committee Member iii

4 DEDICATION To my wife iv

5 A fact is a simple statement that everyone believes. It is innocent, unless found guilty. A hypothesis is a novel suggestion that no one wants to believe. It is guilty, until found effective. (Edward Teller) v

6 AKNOWLEDGEMENTS I would like to thank my advisor, Dr. Hamid Lankarani for his thoughtfulness, patience and comprehension, not only as a professor but also as a person to learn from him. I also thank to members of my committee, Professor George Talia and Dr. Gerardo Olivares, for their helpful comments and suggestions on all stages of this Thesis. I would also like to extend my special gratitude to Dr. Gerardo Olivares who also has been my supervisor at National Institute for Aviation Research (NIAR) during the last years, contributing to my intellectual development and allowing me to utilize the research material. Also thanks to engineer and fellow Tom Berry for his suggestions and text corrections. The support of the Federal Aviation Administration for this research is also highly appreciated. vi

7 ABSTRACT Aircraft regulations for seat certification are adapting to new crashworthiness exigencies. However, the use of the forty year old Hybrid II Anthropomorphic Test Device (ATD) for seat certification has remained invariable in the aircraft community (manufacturers, regulators and researchers) for decades. Although the community has had the improved FAA Hybrid III 50 th ATD (FAA HIII) from more than ten years is still scarcely used for seat certification. Possible reasons of the unpopularity of the FAA HIII might be: (1) concern with rumored more stringent biomechanic responses, which might make it more difficult for seat certification and (2) poor literature available for the FAA HIII in comparison to the Hybrid II ATD (HII). This Thesis deals with a research effort to expand the scarce information available on the FAA Hybrid III 50 th male ATD, focusing in the lumbar-pelvis responses for different aircraft vertical loading conditions. The results from this Thesis research indicate that the FAA Hybrid III 50 th ATD has achieved high degree of repeatability and linearity for lumbar-pelvis responses for Parts 23 and 25 Section 562 Emergency Landing Conditions. The demonstrated FAA HIII s characteristics can help aircraft seat community to predict potential spine responses for different dynamic test configurations and for instance answer what if questions at the beginning of the seat design phase. Thus, the biofidelic improvements of the FAA HIII as well as reliable lumbar-pelvis responses can be sufficient enough reasons for replacing the old Hybrid II for seat certification purposes leading in safer designs. vii

8 TABLE OF CONTENTS Chapter Page 1. INTRODUCTION Motivation Background Principles of Crashworthiness Seat Certification Regulations Injury Criteria Human Spinal Injury Tolerance The FAA HIII 50th Male and Other Aircraft ATDs Literature Review Scope and Objectives of the Research DYNAMIC TESTING METHOD FOR THE FAA HYBRID III 50TH ATD NIAR Sled Facility Description Test Orientation and Pulse Severity Instrumentation Protocol Seat Pan Orientation and Mass Effects Seat Cushion Characteristics RESULTS, DATA ANALYSIS AND DISCUSSION Test Results of the Lumbar-Pelvis Responses of the FAA HIII for 60 o 2 point belt FAA Hybrid III lap-belt only 60 o for Part Rigid Seat FAA Hybrid III lap-belt only 60 o for Part Cushion Seat FAA Hybrid III lap-belt only 60 o for Part Rigid Seat Repeatability Results of the Lumbar-Pelvis Responses of the FAA HIII Selected Methods Repeatability of the FAA HIII for all 60 o Configurations Test Data Analysis and Discussion 48 viii

9 TABLE OF CONTENTS (continued) Chapter Page 4. CONCLUSIONS AND RECOMMENDATIONS Conclusions Recommendations 54 REFERENCES 56 APPENDIX 60 Error Metrics of the FAA Hybrid III ATD 50 th 60 o Configurations 61 ix

10 LIST OF TABLES Table Page 1. Dynamic Seat Requirements for Parts 23, 25, 27 and 29 Section Summary of Injury Pass-fail Criteria per Parts 23, 25, 27, 29 Section ATD General Comparison [6] Dynamic Requirements for 60 Pitch Tests [3] Cushion Characteristics [3] Test Matrix Table for the FAA HIII Horizontal-Vertical Test Configuration Part Rigid Seat [3] Horizontal-Vertical Test Configuration Part Cushion Seat [3] Horizontal-Vertical Test Configuration Part Rigid Seat [3] Evaluation Periods for the FAA HIII 60 o Configurations [3] Measured Repeatability of the FAA HIII for 60 o Configurations [3] FAA HIII Results vs. Correlation Limits for Lumbar Responses [3,21] Comparison of the FAA HIII with Cushion vs. Rigid Seat for Part Repeatability OF the FAA HIII 60 o 2 pt Belt Part Rigid Seat [3] Repeatability OF the FAA HIII 60 o 2 pt Belt Part Rigid Seat [3] Repeatability OF the FAA HIII 60 o 2 pt Belt Part Cushion Seat [3] x

11 LIST OF FIGURES Figure Page 1. High-level requirements map for seat aircraft certification Energy management system for an airplane [8] Test 1 horizontal-vertical crash impact [8,6] Test 2 horizontal crash impact [8,6] Common spine injured zones due aircraft vertical impact forces [8,17] Requirements map for aircraft seat certification FAA HIII ATD lumbar-pelvis modifications [6] Lumbar Fz of the Hybrid II and FAA Hybrid III [6] FAA HIII ATD in a helicopter impact test [18] Numerical model vs. tests results for 60o tests [18] NASA test of a crashworthy fuselage using two HIIs [20] NASA HII s lumbar test results [20] o tests of the FAA HIII conducted at NIAR [3] Rigid seat (left) and cushioned seat for 60 o tests (right) [3] Pulse repeatability obtained at NIAR [3] NIAR rigid seat components [3] Rigid seat orientation w.r.t glob.coord.sys. [3] Cushion stress vs. deflection curve [3] o tests performed at NIAR [3] Side view for Part rigid seat at NIAR [3] Lumbar Fz and My. Part rigid seat [3] Seat pan forces Part rigid seat at NIAR [3] xi

12 LIST OF FIGURES (continued) Figure Page 23. Side view for Part cushioned at NIAR [3] Lumbar Fz and My. Part cushioned [3] Seat Pan Forces Part Cushioned [3] Side view for Part rigid seat at NIAR [3] Lumbar Fz and My. Part rigid seat [3] Seat pan forces Part rigid seat [3] Selected Methods metrics [3] Repeatability of the lumbar forces and Y moments for all configurations [3] Repeatability of the seat pan forces for all configurations [3] Averaged compliance lumbar responses all Parts [3] Lumbar vertical forces for all 60 degrees configurations xii

13 LIST OF ABBREVIATIONS/NOMENCLATURE FAA ATD FAA HIII HIII HII NIAR S&G SM GA HIC CAMI JAR Federal Aviation Administration Anthropomorphic Test Dummy FAA Hybrid III (50 th ) ATD Standard Hybrid III (50 th ) ATD Hybrid II (50 th ) ATD National Institute for Aviation Research Sprague and Geers Metric Selected Methods Metric General Aviation Head Injury Criteria Civil AeroMedical Institute Joint Aviation Rules (European Regulations) xiii

14 CHAPTER 1 INTRODUCTION 1.1 Motivation In October 2009 the final rule for dynamic seats included in 14 CFR Part for transport airplanes was finally implemented after a grace period of more than twenty years [7]. The new requirements add stringent dynamic tests rather than just static tests for seat certification. It is clear that aircraft regulations are evolving with crashworthiness exigencies however the aircraft community still utilizes the forty year old Hybrid II Anthropomorphic Test Device for testing and seat certification purposes. This is not the case of the automotive industry which, in 1991, shifted completely from the Hybrid II to the new generation Standard Hybrid III (HIII) because of improved biofidelity and test measurement capabilities. Similar to the automotive industry, the aircraft community also has had at its disposal for over a decade an improved dummy, the so called FAA HIII [6]. The new aircraft dummy is based on the biofidelic advances of the automotive HIII, but is also adapted to measure vertical impact responses which are required on aircraft regulations. However, contrary to the automotive industries adoption of the FAA HIII for biomechanical testing is infrequently utilized for aircraft seat certification purposes even though regulations allow its use. This Thesis deals with the research efforts to expand the scarce literature available of the FAA HIII 50 th male, focusing in the lumbar-pelvis responses for the combined horizontalvertical dynamic loadings for different test configurations. The analysis of the results collected at National Institute for Aviation Research (NIAR) of the FAA Hybrid HIII 50 th ATD during 1

15 2007 and 2008 demonstrate that the improved dummy has repeatable, stable and linear lumbarpelvis responses. 1.2 Background This Thesis reports the research efforts to expand the scarce information currently available of the FAA Hybrid III 50 th male anthropometric test device (ATD) for seat certification purposes. This work is specifically focused in the lumbar-pelvis responses of the ATD for various combined horizontal-vertical dynamic loadings which are typical in airplane and especially helicopter accidents. Such loads can lead in severe spinal injuries to the human being. The overall purpose of using an anthropometric test device (ATD) is to measure similar responses with an acceptable degree of repeatability and reproducibility that a human being produces when tested for different loading dynamic conditions. Current research has shown that there exists a scarcity of literature regarding the behavior of the FAA HIII for aircraft seat certification, even though it is a new generation ATD with improved biofidelic responses as compared to the widely used but forty year old HII ATD. Possible Rumored causes of the unpopularity of the FAA HIII despite is not written anywhere might be: (1) Concern with reportedly more stringent biomechanic responses in head velocity which might make it more difficult seat certification and (b) scarce literature available when responses are compared with the old HII. The dynamic tests requirements included in the FAA regulations for 14 CFR Parts 25, 23, 27 and 29 section 562 Emergency Landing Dynamic Conditions for aircraft seat certification [11,12,15,16] state that tests must be conducted using an ATD to simulate the aircraft occupant and to measure the injury data. These regulations were initially based on HII responses however, they currently allow the use of both HII and FAA HIII [5,6]. The measures produced on the 2

16 ATD are correlated with four (4) "pass-fail" criteria to determine the degree of injury. Thus, the purpose of all four regulations is to certify seat designs in which the occupant won't be seriously injured during a controlled impact conditions such as an emergency landing. Aircraft seats tested under these regulations must be capable to absorb required impact energy without structural collapse, as well as, protect occupants from being injured from seat component reactions according to defined injury criteria (section 1.2.2). Next figure presents the high-level requirements map to certify aircraft seats. FIGURE 1. High-level requirements map for seat aircraft certification. As previously noted, current regulations allow the use of both the "old" HII and "new" FAA HIII ATD s even thought the last is rarely, if ever, used by the aircraft seat community for certification purposes. However, it is known that the higher degree of bio-fidelity, repeatability and reproducibility of an ATD, the better quality of injury data obtained resulting in safer seat designs. Analysis of the results collected at National Institute for Aviation Research (NIAR) of the FAA Hybrid HIII (50th) ATD during 2007 and 2008 indicate a high degree of repeatability and consistency on all Lumbar-Pelvis responses [3]. The following overview presents key points and definitions that are important to understand the included information. 3

17 1.2.1 Principles of Crashworthiness Aircraft regulations for seat certification are adapting to new crashworthiness exigencies as recently happened with the implementation of dynamic tests in the 14 CFR Part The crashworthiness concept responds to those characteristics that make a vehicle safer in order to better protect the occupants from injuries or fatalities during a crash event [8]. There are five (5) pillars that constitute the crashworthiness discipline; (1) Container, (2) Restraint, (3) Energy management, (4) Environment and (5) Post-crash factors. Container The most important principle refers to the structural integrity of the physical volume that protects the occupants. A strong container is necessary because without this principle all other four pillars are unnecessary [8]. Restraint Once the container is sufficient ly strong to maintain a protective envelope for he occupants the next step is to restrain the occupant within the container. This is the second most important principle when measuring crash worthiness. The main goal of the restraint systems is to maintain the occupant within the container and to transfer inertial loads to the skeletal parts of the occupant s body so as to not harm soft tissues or organs [8]. Energy-Absorbing Management This is a crashworthiness principle that is challenging from an engineering point of view. The goal is to control the peak accelerations and peak loads applied during impacts. 4

18 This principle can be observed implemented in each mechanical subsystem such as seatlanding gear-fuselage in aeroplanes or energy-absorbing bumpers in cars. The next figure represents how fuselage, landing gear and seat collapse during the impact absorbing energy, and thus decreasing the final energy and loading to be absorbed by the occupant [8]. Prior Impact After Impact FIGURE 2. Energy management system for an airplane [8]. Environment The interior design of the vehicle must provide safety to the occupants in case of an impact. For example the substitution of sharp for rounded edges on interior edges and padding of hard interior surfaces would decrease the number of injuries due the impact [8]. Post Crash hazards The last pillar of the crashworthiness discipline is to minimize injuries after the impact is terminated. Avoiding the use of flammable and toxic smoke materials or designing well marked aises and emergency exits are examples included in this principle [8]. 5

19 1.2.2 Seat Certification Regulations At the end of the 1980 s, the FAA increased the degree of exigency on seat aircraft crashworthiness and introduced the dynamic tests requirements for seat aircraft certification [5,7]. The regulations that specify seat aircraft requirements are: (1) 14 Code of Federal Regulation (CFR) Part 23 section for normal, utility, acrobatic, and commuter transport airplanes.(2) 14 CFR Part 25 section for transport category airplanes.(3) 14 CFR Part 27 section for normal Rotorcraft aircraft and (4) 14 CFR Part 29 section for transport category aircraft Emergency Landing Conditions [11,12,15,16]. The scope of all four regulations is to certify seat designs in which the occupant won't be seriously injured during a controlled impact conditions such as an emergency landing. Aircraft seats tested under these regulations must be capable to absorb required impact energy without structural collapse as well as protect occupants from being injured from seat component reactions according to defined injury criteria (section 1.2.3). Impact energy is simulated by applying different deceleration pulses to the specimen. Thus, pulse deceleration severity varies according to the category of the aircraft to certify. For example, smaller General Aviation (GA) aircraft included in Part 23 have stringent pulse severity in comparison any transport airplane included in Part 25, due the structural weakness and less components capable to absorb the energy developed during an impact. All four regulations require two seat test orientations described as (1) test 1 and (2) test 2. Table 1 summarizes the dynamic test requirements for all 14 CFR Parts. 6

20 Test 1 Test 1 is a combined horizontal-vertical orientation also known for seat aircraft community as the 60 pitch tests when those are conducted on horizontal sleds. This orientation simulates an aircraft crash with the longitudinal axis orientated 30 nosedown with respect to the crash plane therefore the resultant impact force has a significant vertical component. This test is required to evaluate the forces and moments applied on the spine column of the occupant. FIGURE 3. Test 1 horizontal-vertical crash impact [8,6]. Test 2 Test 2 is only horizontal seat orientation. This test simulates an aircraft crash with only horizontal inertial forces developed to the occupants. This type of test is important to determine responses, accelerations and excursions of the principal human body regions as well as defining interior part locations to prevent injuries from secondary collisions. 7

21 FIGURE 4. Test 2 horizontal crash impact [8,6]. TABLE 1. DYNAMIC SEAT REQUIREMENTS FOR PARTS 23, 25, 27 AND 29 SECTION 562. DYNAMIC TEST REQUIREMENT [11,12,15,16] PART Utility airplane PART Transport airplane PART Normal rotorcraft PART Transport rotorcraft TEST 1 (Horizontal-Vertical) Test Velocity- Ft/Sec Seat Pitch Angle Deg. Seat Yaw Angle Deg. Peak Acceleration- G s Time To Peak Sec Floor Deformation Deg / /0.06 None None Pitch /10 Roll Pitch /10 Roll TEST 2 (Horizontal) Test Velocity- Ft/Sec Seat Pitch Angle Deg. Seat Yaw Angle Deg. Peak Acceleration- G s Time To Peak Sec Floor Deformation Deg ±10 26/ / Pitch /10 Roll 44 0 ± Pitch /10 Roll 42 0 ± Pitch /10 Roll 42 0 ± Pitch /10 Roll As is noted at the beginning of this section the resulting ATD response must be correlated with various pass-fail injury criteria to make sure that the occupant is not seriously injured at the prescribed loadings and test conditions. 8

22 1.2.3 Injury Criteria Injury criterion is the baseline specification used to determine whether dynamic responses such as acceleration of body segments or internal loads recorded on an Anthropomorphic test Device (ATD) are harmful for the human being. FAA specifies four (4) injury criteria to use for seat certification purposes; (1) Belt load criterion, (2) Head injury criterion (HIC), (3) Femur compressive load criterion and (4) Spinal compressive load criterion. Similarly to the seat test regulations the injury criterion have differences among regulations according to the aircraft category [5,11,12,15,16]. Table 2 summarizes all allowable maximum values for each criterion and Part. Belt Load Criterion The belt load criterion focuses on the tension allowed on the belt straps. The purpose of this criterion is to avoid injuries in the thoracic body region due the belt loads when tensioned in an impact. There are (3) three key points defined in this criterion [5,11,12,15,16]: When no shoulder strap is provided (lap-belt-only), as is the case of passengers in transport airplanes, no belt tension limit is required. If there is one shoulder strap (three-point-belt) typical restraint systems for pilots in GA, the minimum total tension load allowed is 1,750 lbs. The last case is if the belt is equipped with two shoulder straps (four-point-belt) which is a restraint system commonly used for pilots in transport category airplanes. The total tension allowed must be less than 2,000 lbs (the sum of forces due both straps). 9

23 Head Injury Criterion (HIC) HIC is based in the relation with head transitional acceleration and time duration. High head accelerations developed in short periods of time will cause injury. However, if the head acceleration is lower the degree of severity may decrease [1,5]. For aviation certification purposes, HIC is calculated when head impacts occur and only for the impact period of time. HIC injury measurements above 1,000 (unit-less) are considered severe head injuries whereas for HIC values below 1000 the probability of injury becomes much less significant [5,11,12,15,16]. HIC ( t 2 t 1 ) ( t 2 1 t 1 ) t2 t1 a( t) dt Where: t 2,t 1: time collision limits which a(t) reaches its peak value expressed in msec. a(t): Translational acceleration of the head center of gravity expressed in G s. 2.5 Femur compressive Load Criterion This criterion measures the compressive load applied on each femur to determine injuries in the lower body region. Forces (each femur) above 2,250 lbs are considered harmful to humans [5,11,12,15,16]. Spinal Compressive Load Criterion The spinal criterion specifies a compressive maximum load on the ATD s between the lumbar and pelvis region of 1,500 lbs which corresponds a 20% risk of 10

24 moderate spine human injury according to the Abbreviated Injury Scale (AIS 2+) [19]. Researchers [19] derived a quasi-static lumbar spine compression force limit of 1,500 lb. based on the Dynamic Response Index (DRI) which is a simplified mass-spring-damper model to simulate the biomechanical responses due the impact of ejection seat to the human body. Most critical spine injuries are produced due to direct impact during the structural deformation of the seat system causing loading to this region. Same researchers used the DRIz (DRI for the vertical direction) value of 19 which corresponds for operational data: 9% injury risk AIS 2+; and for laboratory data: 20% injury risk AIS 2+ (moderate spine injury) when using a Hybrid II with straight spine column. TABLE 2. SUMMARY OF INJURY PASS-FAIL CRITERIA PER PARTS 23, 25, 27, 29 SECTION 562. Injury Criteria [11,12,15,16] PART Utility airplane PART Transport airplane PART Normal rotorcraft PART Transport rotorcraft Total Belt strap loads (lbs) (one strap/two straps) 1,750/2,000 1,750/2,000 1,750/2,000 1,750/2,000 Spinal Compressive load (lbs) 1,500 1,500 1,500 1,500 Femur load (lbs) Not required 2,250 Not required Not required HIC 1,000 1,000 1,000 1,000 Note that lumbar-pelvis results of the FAA HIII (chapter 3) must provide suitable biofidelic response to be able to correlate with the spinal compressive criterion. To better understand this criterion an overview of spine injury tolerance is next introduced. 11

25 1.2.4 Human Spinal Injury Tolerance The spinal column is one of the vulnerable parts of aircraft occupants in case of an accident. These types of impacts are common in airplane and especially helicopter accidents where there is a vertical component of the crash which would be for most accidents [9]. These injuries are frequently located on the upper lumbar and lower thoracic regions (T10-L2) and may result in paralysis or other neurological symptoms [17,8]. Figure 5 below show the anatomy of the spinal column and the most common spinal injured zone due vertical impact forces (highlighted in red). FIGURE 5. Common spine injured zones due aircraft vertical impact forces [8,17]. It is demonstrated that the spinal failure is a sum of damages in the soft and hard tissues [8]. In addition sensitive changes on the impact scenarios may quickly result in different degrees of spinal injury. For example small differences on the eccentricity of the applied impact force, as well as small variations of the initial and end position of the column can change the degree of injury[8,9]. The variable injury characteristics and also lack of literature available in comparison with other body regions makes difficult to develop a precise injury criteria for the spine column. As is 12

26 noted before, current FAA spinal criterion refers to the maximum compression load (1500 lbs) between the lumbar spine and the pelvis [11,12,15,16] However, pure compression loading is unusual and usually the injury is produced from combination of various load mechanisms [8]. Next are defined the four (4) basic load mechanisms attributed to the spine column and type of injuries, which are classified according to the primary direction of loading [8]. Axial Compression This type of load mechanism produces fracture-type injuries to the vertebral bodies of the column, especially in zone T10-L2. When compression is accompanied with flexion the injuries produced are anterior wedge and burst fractures [8]. Axial Rotation/Torsion Frequently combined with other load mechanisms such as compression or shear loads and cause severe neurological injuries and paraplegia [8]. Shear Usually this load mechanism combines with flexion and rotation mechanisms resulting in impingement of the column difficult to repair surgically [8]. Spinal Flexion Usually spinal flexion combines with all other four mechanisms. When combined with other loadings flexion may produce unilateral and bilateral dislocations, anterior wedge and burst fractures [8]. Spinal flexion is also common when the submarining 13

27 occurs. This phenomenon can be produced during vertical-horizontal impact forces where the occupant slips under the belt while the column flexes over it. This phenomena produces severe spinal column injuries due the pressure applied on spine rather than the pelvis resulting in the separation of the posterior components of the vertical bodies [8,9]. In conclusion, the spinal compressive criterion is currently the accepted way during testing with an ATD to determine potential degree of injury for pure vertical loads. However, spinal column injuries as noted in this section can be produced from a combination of various load mechanisms rather than just compression loading. Figure 6 schematizes the concepts defined in this chapter and the correlations among them for a current seat certification process. FIGURE 6. Requirements map for aircraft seat certification. 14

28 1.2.5 The FAA HIII 50 th Male and Other Aircraft ATDs An anthropomorphic test device (ATD) also called dummy is a mechanical device designed to measure similar responses a that human being produces when tested under different loading and dynamic conditions [5,8]. Dummies must be capable to meet (5) five requirements [5,8]: Bio-fidelity and Anthropometry How the ATD represents the physical characteristics of the human being; for example body measures, mass distribution, body inertias, stiffness, and flexibility.[8] Reproducibility How well two identical ATDs generate the same response when tested under same conditions [5]. Repeatability The capability that an individual ATD to respond equally when conducting different tests under same initial conditions. Durability or replacing parts. The ATD has to maintain its integrity during various tests before calibrating 15

29 Measurement accuracy The ATD as an instrument device must be able to be calibrated periodically in aim to obtain the maximum accuracy on the measures. The dynamic tests requirements for 14 CFR Parts , , and Emergency Landing Dynamic Conditions for aircraft seat certification state that tests must be conducted by using an ATD [11,12,15,16]. The measurements recorded with the ATD are compared with the pass/fail criteria to determine the degree of potential injury These FAA dynamic test requirements were developed during the 1980 s based on the automotive HII 50 th male ATD which is still being widely used in the biomechanics discipline particularly for aircraft testing. Originally, The HII was introduced in 1972 by General Motors in aim to evaluate the behavior of the lap and shoulder belt restraint systems in frontal impacts for the automotive industry and represented the anthropometric average for the fifty per cent (50%) of the U.S. male population. The HII demonstrated adequate repeatability and reproducibility [6]. A next generation of ATD, the Hybrid III 50 th male ATD was developed for the automotive research and regulations for frontal impact dynamic tests. The major improvements obtained were; a more detailed neck, head, thorax, and pelvis lumbar region. In addition the HIII allowed to gather more biomechanical data and had better biofidelity. In 1986 the National Highway Traffics Safety Administration (NHTSA) included the HIII in the automotive regulations [6] as a valid test Dummy, and within five years (1991) the HIII totally replaced the HII for occupant crash automotive regulations [8] However, for aircraft certification purposes especially for the combined horizontal-vertical tests (tests 1) and associated lumbar criteria, the new automotive standard HIII dummy showed 16

30 design incompatibilities in the lumbar and pelvic regions and was not accepted by the aircraft authorities as a valid aircraft test device[6]. The automotive standard HIII has the spine column curved which replicates a more slouched seated position typical in a car and was not designed to measure vertical loads. In comparison, the HII has a straight lumbar column resulting in an erect-spine seated posture that better replicates the aircraft seated position and for being capable to measure vertical impacts. In addition, the mass in the pelvis area in the automotive HIII was fifty per cent heavier (50%) than the HII affecting the upper torso and head kinematics [6]. The design differences between the aircraft HII and the automotive standard HIII were enough to generate unacceptable differences in key impact responses required for seat aircraft certification such as the lumbar vertical force (Fz) and head excursion [6]. These differences did not allow the standard HIII to be accepted as a valid test device for aircraft seat certification. Because aircraft community desired to introduce the biofidelic advantages of the standard HIII such as neck, head, and lower legs for research and seat certification, a group of seat aircraft experts developed the FAA HIII 50 th male based on the biofidelic improvements and measuring capabilities of the automotive standard HIII adapting the lumbar and pelvis regions for aircraft regulations [6]. The FAA Hybrid III 50 th ATD The FAA Hybrid III 50th Percentile Male ATD is a dummy developed by FAA, Civil AeroMedical Institute (CAMI), Denton ATD, Inc., and Robert A. Denton, Inc., for Federal Aviation seat certifications. 17

31 The primary design goal was to take advantage of the biofidelic improvements of the automotive standard Hybrid III while being capable to replicate all impact responses equivalent to those of the Hybrid II [6]. The FAA HIII is constructed by using standard components from both the Hybrid II and Standard Hybrid III ATD s to minimize new designs needed to configure the straight-spine of the HII. To take advantage of the biomechanics improvements developed on the Std. HIII, The FAA ATD utilizes the head, neck, thorax and lower limbs of the automotive dummy [6,13]. The six (6) modifications performed on the Std HIII to adapt the pelvis-lumbar region are: [6] 1). Substitution of the Hybrid III lumbar-pelvic adapter block with Denton Model 1891 Hybrid II Lumbar Load Cell and its pelvic adapter block to record lumbar responses [6] 2).Replace the curved Hybrid III lumbar column with a the HII s straight column, which attaches to the Denton lumbar load cell and pelvic adapter block assembly.[6] 3). Substitute the Hybrid III upper lumbar-thorax adapter with a new adapter made in steel rather in aluminium to closely reproduce the mass distribution of the HII s upper torso.[6] 4). Substitution of the Hybrid III abdominal insert for a standard Hybrid II abdominal insert [6]. 18

32 5). Replacement the Hybrid III upper leg boy parts for the Hybrid II upper leg body parts [6]. 6). Replacement the Hybrid III chest flesh jacket with the chest flesh from a Hybrid II conveniently adapted to allow neck to move free [6]. Figure 7 shows all the modifications performed on the FAA HIII 50 th male ATD. FIGURE 7. FAA HIII ATD lumbar-pelvis modifications [6]. Unfortunately, there is not much literature available for the FAA HIII. This was one of the reasons to write and report this Thesis. Frequently papers written by seat aircraft community usually refer to and utilize the old HII. Table 3 shows the ATD comparison for seating postures, sketches of their lumbar-pelvic regions and basic measurements; (1) HII, (2) automotive Std. HIII and (3) FAA HIII. 19

33 TABLE 3. ATD GENERAL COMPARISON [6]. ATD Seating Postures Comparison HYBRID II 50 th male FAA Hybrid III 50 th male Std. Hybrid III 50 th male Upper Torso (lbs) 57.2 Upper Torso (lbs) 55.0 Upper Torso (lbs) 55.6 Arms (lbs) 19.1 Arms (lbs) 19.4 Arms (lbs) 19.4 Lowe Torso (lbs) 35.4 Lowe Torso (lbs) 33.6 Lowe Torso (lbs) 45.4 Upper Legs (lbs) 35 Upper Legs (lbs) 33.4 Upper Legs (lbs) 26.2 Lower Legs (lbs) 19.8 Lower Legs (lbs) 25.3 Lower Legs (lbs) 25.3 Total Weight (lbs) (Specified) 164.0±3 Total Weight (lbs) (Specified) 164.0±3 Total Weight (lbs) (Specified) 172.3±2.4 Sitting Height (in) (Specified) 37.7±0.1 Sitting Height (in) (Specified) 37.7±0.1 Sitting Height (in) (Specified) 34.8±0.2 20

34 1.3 Literature Review Van Gowdy et al. A Lumbar Spine Modification to the Hybrid III ATD for Aircraft Seat Tests [6] is still used as the reference document related to this matter although was published ten years ago (1999). Researchers conducted several tests to compare compliance responses of the FAA HIII, the Std. HIII and HII for Part Transport airplanes only which has the less severe test conditions in comparison with other 14 CFR Parts. Figure 8 shows the results for 60 o lap-belt only with 14 G s pulse and rigid seat. In these tests the HII and FAA HIII passed the spinal compression criteria with similar behaviour. FIGURE 8. Lumbar Fz of the Hybrid II and FAA Hybrid III [6]. Although other testing was out of the scope of Van Gowdy et al, it was noted that testing the ATDs for more severe conditions, as for example those required in Parts 27 and 29 for rotorcraft, other ATD differences might be observed. Researchers concluded that the FAA HIII compliance responses were close to the HII although slightly larger differences in the head velocity would result in higher HIC results and the authors stated that might be considered a conservative approach. 21

35 Boucher and Waagmeester [18] published in 2003, Enhanced FAA Hybrid III Numerical Dummy Model in MADYMO for aircraft Occupant safety assessment. This work among others was included under the European project HeliSafe 1. Researchers studied compliance responses of the physical FAA HIII ATD and compared with a numerical model modelled in MADYMO for helicopter crash events. Combined horizontal-vertical lap-belt only tests were conducted according to Parts 27 and 29 for rotorcraft with a 30 G s pulse. Although the primary goal of the research was to validate the numerical model as a certification tool, Helisafe chose the FAA HIII rather than HII to conduct the tests because; (1) the dummy is the most advanced in its kind, (2) It duplicates the Hybrid II performance in dynamic seat testing required by FAA and JAA, (3) It is accepted by the FAA as alternative to the HII specified in FAR/JAR 23,25,27 and 29 Section 562, and (4) The FAA HIII part will be available for maintenance for at least 10 years. In addition to these issues, Helisafe opted for the FAA HIII due the use of airbags and advanced safety features. The use of such safety issues is a current crashworthiness tendency for new helicopters and airplane designs, especially in General Aviation. The FAA HIII improvements allow recording head and neck data in impact events in which the airbag is used, whereas HII does not. Experimental results for test S are shown in Figure 10. Note that the lumbar load passed the criteria obtaining values between 5500 N (1236 lbf.) for the upper spine and 6250 N (1405 lbf.) for the lower spine. Unfortunately the paper presents scarce test data even though the researchers performed four 60 o tests with 30 G s pulses. 1 Helisafe project TM : Helisafe is a European Commission founded project whose aim is to improve chances of helicopter cockpit ad cabin occupants surviving crash and to reduce the risk of injuries. The target is to reduce crash fatalities and major injuries by 50%, a figure that will be achieved using an advanced Cabin Safety System concept based on interactive safety features. 22

36 FIGURE 9. FAA HIII ATD in a helicopter impact test [18]. FIGURE 10. Numerical model vs. tests results for 60 o tests [18]. Fasanella and Jackson 2002 in Impact Testing and Simulation of a Crashworthy Composite Fuselage Section with Energy-Absorbing Seats and Dummies [20], performed a 25- ft/s vertical drop test of a composite airplane fuselage section. They used two HIIs to determine the interaction between the crashworthy characteristics of the fuselage and ATDs seated on energy absorbing seats. U.S. Army researchers correlated the test results with the predictions of a crash numerical model developed with MSC. Dytran. The results of the lumbar load were 23

37 different for both dummies. Although both dummies were close to 1500 lbs. one passed the criteria (seated on the right side) and the other failed (left side). Unfortunately this study was not performed by using FAA HIII s even though the research was conducted in Next figures 11 and 12 show the pre-test, the post-test, and experimental results of the lumbar vertical load obtained for both HIIs after the impact. FIGURE 11. NASA test of a crashworthy fuselage using two HIIs [20]. FIGURE 12. NASA HII s lumbar test results [20]. 24

38 At the moment of writing this Thesis, Olivares Dynamic Seat Certification by Analysis: Volume I, II and III [2,3,4] is concluding an extensive study sponsored by the FAA of the HII and FAA HIII. The NIAR researcher has conducted during two years a full matrix of tests for Part 25 and 23 Section 562 using both ATDs. The primary goal of the research is to develop a measurable tool for the FAA to determine whether the quality and methods used in the industry for testing and model aircraft restraint systems are appropriate. Because of the amount of data collected for both hybrids, a comparison of the compliance responses of the HII and FAA HIII will be studied as well. Unfortunately, the complete results and final conclusions from this extensive study are currently awaiting publication. Results shown in this paper of the lumbarpelvis region of the FAA HIII are collected entirely from the Olivares and FAA project [3]. FIGURE o tests of the FAA HIII conducted at NIAR [3]. 1.4 Scope and Objectives of the Research The purpose of this research is to present compliance lumbar-pelvis data of the FAA Hybrid III 50th for 14 CFR Part 23 and 25 section 562 Emergency Landing Conditions for combined horizontal-vertical dynamic loadings. This research will deal with the analysis of data collected from a number of dynamic tests conducted at NIAR during the period of

39 The eventual goal of this research is to persuade aircraft community of the measurement capabilities of the FAA HIII for combined horizontal-vertical dynamic loadings. Furthermore, other dummy improvements such as head, neck and lower leg regions (out of the scope of this work) with respect to the HII will result in more biofidelic approaches for compliance responses that will improve aircraft safety. In general, the present research is expected to be of great use to the airplanes and rotorcraft manufacturers, aircraft seat manufacturers, the FAA and JAR authorities, numerical modeling companies and aircraft researchers. 26

40 CHAPTER 2 DYNAMIC TESTING METHOD FOR THE FAA HYBRID III 50 TH ATD During several dynamic seat tests were conducted at NIAR of the FAA HIII 50 th male to study its body responses for Part and Emergency Landing Conditions. These tests are included in Olivares and FAA project Dynamic Seat Certification by Analysis: Volume I, II and III [2,3,4] who is concluding an extensive study sponsored by the FAA of the HII and FAA HIII. The primary goal of the research is to develop a measurable tool for the FAA to determine whether the quality and methods used in the industry for testing and model aircraft restraint systems are appropriate. This section presents an overview of the testing methodology used for horizontal-vertical loadings lap-belt-only commonly referred 60 pitch test. The same orientation tests for Part and Part were conducted by using a rigid seat to obtain pure compression responses on the dummy and seat. The reason to perform tests with a rigid structure rather than a real aviation seat was to avoid interferences such as non-desirable deformations or dynamic stiffness. In addition, for only Part tests were also conducted with a non flotation aircraft seat cushion [3] to determine baseline differences with only rigid seat tests. FIGURE 14. Rigid seat (left) and cushioned seat for 60 o tests (right) [3]. 27

41 2.1 NIAR Sled Facility Description NIAR sled facilities are located in Wichita State University in Wichita, Kansas at the Crash Dynamics Lab. This installation has more than 4000 square feet with controlled room temperature. The Lab conducts the dynamic tests by using a Servo-hydraulic crash simulator, Model with an accelerator sled which has a high degree of repeatability and accuracy to replicate almost any type of crash pulse. The facility is used for research, testing and certification of aircraft and automotive components [3]. Figure 15 shows the high degree of pulse repeatability that the NIAR crash simulator and sled are able to obtain among 6 tests with identical set up configuration. FIGURE 15. Pulse repeatability obtained at NIAR [3] 2.2 Test Orientation and Pulse Severity NIAR tests have been rigorously conducted as is required in CFR Parts 23 and 25 section 562 [11, 12] for 60 pitch tests. According to both CFR Parts there are to different severity pulses: For (1) Part the pulse to apply simulates a variation in the downward velocity of no less than 35 ft/sc with the airplane's longitudinal axis canted downward 30 with respect to the horizontal plane, and with the wings leveled. Peak floor deceleration must occur in not more 28

42 than 0.08 sec after impact and must reach a minimum of 14 G s [11]. Whereas for (2) Part as is stated Change in velocity may not be less than 31 feet per second. The seat/restraint system must be oriented in its nominal position with respect to the airplane and with the horizontal plane of the airplane pitched up 60 degrees, with no yaw, relative to the impact vector Peak deceleration must occur in not more than 0.05 seconds after impact and must reach a minimum of 19g (more severe conditions which corresponds to the pilot seat) [12]. Table 4 summarizes the dynamic test requirements specified in Parts and conducted at NIAR for the FAA HIII. TABLE 4. DYNAMIC REQUIREMENTS FOR 60 PITCH TESTS [3]. DYNAMIC REQUIREMENTS FOR TEST 1 HORIZONTAL-VERTICAL ORIENTATION PART [12] PART [11] Test Velocity- Ft/Sec Seat Pitch Angle Degrees Seat Yaw Angle Degrees Peak Acceleration- G s Time To Peak Sec Instrumentation As is specified in Olivares Work [3], test data is gathered by a Transient and Signal Acquisition/Processing System from DSP Technology, Model VX2850B. This system is capable of recording 96 Channels with a programmable sampling rate up to 50K samples/sec/channel; programmable analog filtering; 16-bit resolution; simultaneous sampling; digital filters, flat pass band, sharp cutoff; pre- and post-trigger recording. The Software used to process the data also defined for the same author is a DSP Technology IMPAX-SD Version 6.0 Data Collection Control and Processing Software. The post processing digitally filters the raw data to the 29

43 appropriate channel class filter in accordance with SAE J211, , and AS8049, [3]. All recording sensors installed on the ATD are typical on a standard FAA HIII 50 th male device. Focusing to the lumbar-pelvis region, the load cell used to record force data is a Denton Model 1891 [3] which is a common one for aircraft seat certification tests. Lap belt is recorded with tri-axial load cell whereas the Seat pan is a six channel Denton 2513 load cell with a 10,000 lbs load and 25,000 lb-in moment capacity in all three axes [3]. Both are mounted on the rigid seat as is shown in figure below. FIGURE 16. NIAR rigid seat components [3]. 2.4 Protocol NIAR researchers conducted at least (3) three identical impact tests for the three (3) 60 configurations to obtain enough data to determine the dummy s repeatability. Engineers used the same dummy (serial number 289) for all impact tests except on the cushioned tests which was also used the dummy serial number 290 for one test. There were (4) four common steps defined for each test [3]: 30

44 Pre-test setup Pre-test setup consists in placing together the rigid seat, the ATD, the load cells and accelerometers. Although the kinematical study is out of the scope of this work, several motion targets were also identified on the ATD relative to the seat reference point to record the motion tracking of each body part. All of these points were conveniently registered in a test-pre data sheet as well as the ATD head and torso 2D profiles as is documented in Olivares work [3]. Test Run and Data Collection This is the phase in which dynamic pulse (14g or 19g) was applied according to the CFR Part tested. Sensors recorded the data during approximately 300 msec at 50 khz. Post-test procedures During this phase the dynamic and load data were checked to determine whether all was conveniently recorded as well as after impact body locations were also documented. Test Data Processing and Channel filters Finally, all data recorded by the accelerometers and load cells were amplified, filtered and digitalized per each channel according to what is specified in SAE (The Engineering Society for Advancing Mobility Land Sea Air and Space ) [3,14]. 31

45 2.5 Seat Pan Orientation and Mass Effects There are two important points for the Seat pan of the rigid seat, first (1) the orientation with respect to (w.r.t) the global axis and second (2) its mass effects on the ATD responses. The seat pan and the seat back (out of scope in this work) are rotated with respect to the global coordinate system. Therefore, loads and moments recorded in the load cells need to be reorientated to the global coordinate system. The seat back plane forms an angle of 13º with the vertical global axis and Seat Pan forms an angle of 5.2º with respect to the horizontal global axis. Figure below show the orientation of the rigid seat respect to the global coordinate system. FIGURE 17. Rigid seat orientation w.r.t glob.coord.sys. [3]. Real seat pan loads and load cell inertia were previously obtained on two only sled rigid seat tests.. In order to obtain pure ATD lumbar responses the values recorded were adjusted by subtracting the seat pan load and load cell inertia obtained from only sled tests,. Furthermore, the weight of the cushion for part cushioned tests was also considered into the adjustment. 2.6 Seat Cushion Characteristics Cushioned tests for Part were also performed during Olivares research [3]. The researcher used the same material properties specified in Development and Validation of an 32

46 Aircraft Seat Cushion Component Test-Volume I [22]. Table 5 characteristics and stress deflection curve (Figure 18) are presented below. TABLE 5. CUSHION CHARACTERISTICS [3]. TYPE LONG (in.) WIDE (in.) THICK (in.) MATERIAL PROPERTIES [18] DENSITY (lb/in. 3 ) Non-flot NAwa3R450/NAwa3C450 With Circ. Coupons of 7.5 in. diam. 4.3 FIGURE 18. Cushion stress vs. deflection curve [3]. 33

47 CHAPTER 3 RESULTS, DATA ANALYSIS AND DISCUSSION This chapter presents and discuses the test and repeatability data of the compliance lumbarpelvis responses and seat pan reactions of the FAA Hybrid III 50 th male ATD for 60 o dynamic tests defined in Parts 23 and 25 sections 562 lap-belt-only. Tests were included in Dynamic Seat Certification by Analysis: Volume II FAA Hybrid III ATD Dynamic Evaluation NIAR Test Series. [3]. Tests were performed at National Institute for Aviation Research (NIAR) during 2007 and The information presented in this chapter is divided in three parts: the first part presents the test results, the second part presents the repeatability results, and the third part analyses and discusses the results. 3.1 Test Results of the Lumbar-Pelvis Responses of the FAA HIII for 60 o 2 Point Belt A total of nine 60 o lap-belt only tests among other seat and belt configurations were performed at NIAR to determine the all ATD body responses as well as hardware reactions (Seat back and Seat Pan). For the 60 o lap-belt only tests there were three configurations and each configuration was repeated three times (nine tests) to obtain enough data to assess repeatability. The three configurations and tests numbers for the 60 o tests were: (1) Part G s no cushion/ tests , -8, and -28, (2) Part G s cushion/ tests , -22, and -23, and (3) Part G s no cushion/ tests , -14, and -15. Although this thesis research is focused in the Lumbar vertical force, another channels related with it are also included such as seat pan reactions or Lumbar Fx and My responses. For each configuration data is presented in form of plots and the initial test set up is presented in form of tables. Plots 34

48 present the data for each single channel for each configuration and tables summarize testing set up information, dummy serial number and peak lumbar vertical (Fz) force results. The data included in each plot allowed to study the repeatability presented in section 3.2 of this chapter. Time was trimmed from the beginning of the pulse t=0 to 300 ms. Figure 19 summarizes the 60 o tests and conditions performed at NIAR [3]. FIGURE o tests performed at NIAR [3]. Table 6 shows the test matrix for the nine tests performed at NIAR. Also included for quick reference are the peak lumbar load obtained for each test as well as the average Lumbar load for each configuration (the values highlighted in red were above the spinal compression criteria). 35

49 TABLE 6. TEST MATRIX TABLE FOR THE FAA HIII. TEST #. ATD Serial# ANGLE (deg) LOAD. SEAT TYPE CRASH PULSE LUMBAR LOAD (lbf) AVER. PEAK LUMBAR Fz (lbf) SPINE CRIT. (lbf) FAA HIII 289 FAA HIII 289 FAA HIII 289 FAA HIII 289 FAA HIII 289 FAA HIII 289 FAA HIII 289 FAA HIII 289 FAA HIII g Rigid g Rigid g Rigid g Rigid g Rigid g Rigid g Cushion g Cushion g Cushion FAA Hybrid III Lap-Belt Only 60 o for Part Rigid Seat The FAA HIII lap-belt only 60 o for Part Rigid Seat include tests numbers: , -8, and -28. Tests were conducted based on same conditions to asset repeatability. The configuration was two point belt (100% Nylon) with seat orientated at 60 degrees from the crash vector. The seat was Rigid and tests were conducted under specifications defined in Part for transport airplanes. The dynamic tests were performed with a 14 G s. Table 7 shows the initial dynamic setup and peak lumbar vertical force for reference. Figure 20, 21 and 22 show the FAA Hybrid III ATD at the beginning of the pulse (t=0) and in some moment after the peak Lumbar vertical force is reached. 36

50 TABLE 7. HORIZONTAL-VERTICAL TEST CONFIGURATION PART RIGID SEAT [3]. TEST #. ATD Serial# ANGLE (deg) LOAD. SEAT TYPE CRASH PULSE LUMBAR LOAD lbf (Compression) FAA HIII 289 FAA HIII 289 FAA HIII g Rigid g Rigid g Rigid FIGURE 20. Side view for Part rigid seat at NIAR [3]. FIGURE 21. Lumbar Fz and My. Part rigid seat [3]. 37

51 FIGURE 22. Seat pan forces Part rigid seat at NIAR [3] FAA Hybrid III lap-belt only 60 o for Part Cushion Seat The FAA HIII lap-belt only 60 o for Part Rigid Seat include tests numbers: , -22, and -23. Tests were conducted based on same conditions to asset repeatability. The configuration was a two point belt (100% Nylon) with seat orientated at 60 degrees from the crash vector. The seat was rigid with a non-flotation cushion and tests were conducted under specifications defined in Part for transport airplanes. The dynamic tests were performed with a 14 G s. Table 8 shows the initial dynamic setup and peak lumbar vertical force for reference. Figures 23, 24 and 25 show the FAA Hybrid III ATD at the beginning of the pulse (t=0) and in some moment after the peak Lumbar vertical force is reached. 38

52 TABLE 8. HORIZONTAL-VERTICAL TEST CONFIGURATION PART CUSHION SEAT [3]. TEST #. ATD Serial# ANGLE (deg) LOAD. SEAT TYPE CRASH PULSE LUMBAR LOAD lbf (Compression) FAA HYB III 289 FAA HYB III 289 FAA HYB III g Cushioned g Cushioned g Cushioned FIGURE 23. Side view for Part cushioned at NIAR [3]. FIGURE 24. Lumbar Fz and My. Part cushioned [3]. 39

53 FIGURE 25. Seat Pan Forces Part cushioned [3] FAA Hybrid III lap-belt only 60 o for Part Rigid Seat The FAA HIII lap-belt only 60 o for Part Rigid Seat include tests numbers: , -14, and -15. Tests were conducted based on same conditions to asset repeatability. The configuration was two point belt (100% Nylon) with seat orientated at 60 degrees from the crash vector. The seat was rigid and tests were conducted under specifications defined in Part for transport airplanes. The dynamic tests were performed with a 19 G s. Table 9 shows the initial dynamic setup and peak lumbar vertical force for reference. Figures 26, 27 and 28 show the FAA Hybrid III ATD at the beginning of the pulse (t=0) and in some moment after the peak Lumbar vertical force is reached. 40

54 TABLE 9. HORIZONTAL-VERTICAL TEST CONFIGURATION PART RIGID SEAT [3]. TEST # ATD Serial # ANGLE (deg) LOAD. SEAT TYPE CRASH PULSE LUMBAR LOAD lbf (Compression) FAA HYB III g Rigid FAA HYB III g Rigid FAA HYB III g Rigid FIGURE 26. Side view for Part rigid seat at NIAR [3]. FIGURE 27. Lumbar Fz and My. Part rigid seat [3]. 41

55 FIGURE 28. Seat pan forces Part rigid seat [3]. 3.2 Repeatability Results of the Lumbar-Pelvis Responses of the FAA HIII As is described in previous chapters a test dummy must accomplish various requirements to be a useful device for test requirements [5]. Repeatability and Reproducibility (section 1.2.5) are the basis of those requirements according to the quality of the output data. This section focuses in the analytical study of the repeatability or variability which is the capability that an individual ATD (serial number 289) to respond equally when conducting different tests under same initial conditions. These responses for the FAA HIII were those collected at NIAR [3] and the repeatability is measured using Selected Methods metrics (SM). 42

56 SM has been chosen by experts in the aviation seating community (SAE aircraft committee working group ARP 5765 [21]) the most adequate metric method to determine the discrepancies between the same responses collected from different aviation seat tests [3,10]. To obtain acceptable ATD repeatability final results should be below the minimum allowable limits of correlation defined by those experts [3,10] Selected Methods Selected Methods is a combined metrics to quantify the repeatability or variability among two curves f(t) and g(t). This method is a combination of Sprangue and Geers (S&D) and Relative Error metrics [3]. The results are expressed if form of three errors, which are; magnitude, delta and shape. Magnitude error It is expressed in percentage (%) and is the difference between peaks of both responses (either positive or negative) biased towards one of the two responses analyzed (reference)[3]. Magnitude error = Maxg( t) Maxf ( t) Maxf ( t) Where: Max f(t) = Peak value positive either negative in reference data. Max g(t) = Peak value positive either negative in data to compare. Delta error It is expressed according to the units of the magnitudes to compare; lbs, lbs in, etc. This error defines the difference (positive) between both peaks thus is not biased [3]. 43

57 Delta error = Maxg( t) Maxf ( t) Shape error It is also expressed in percentage (%) and is the error from S&D metrics which considers both; a magnitude error (not sensitive to phase differences) and a phase error (not sensitive to peak magnitude differences)[3]. This method is biased towards the reference data f(t). (see figure 29) Shape Error = 2 M P 2 where, M = magnitude error factor. P = phase error factor. I gg M = 1 I ff 1 P = * ar cos I I ff fg * I gg Where, I ff = t 2 t2 1 2 t 1 * t 1 f ( t) dt I fg = t 2 1 t 1 * t t 2 1 I gg = t 2 t2 1 2 t 1 f ( t) * g( t) dt * t 1 g ( t) dt Where, t 1 <t<t 2 Time evaluation period. f(t) = Reference data. g(t) = Data to compare. 44

58 FIGURE 29. Selected Methods metrics [3] Repeatability of the FAA HIII for all 60 o Configurations This section presents the repeatability results obtained from Dynamic Seat Certification by Analysis: Volume II FAA Hybrid III ATD Dynamic Evaluation NIAR Test Series. [3] Technical Report FAA-002D, Wichita, KS, September Using Selected Methods described in previous section of the Lumbar-Pelvis responses of the FAA Hybrid III. The results here presented are obtained from work by Olivares [3]. Repeatability is measured in form of percentage error (Magnitude and Shape) and in response units error (Delta). Responses were compared for a specific evaluation period [3]. 45

59 Select Methods considers only the time in which the response is considered important. For example for the Lumbar Fz is important to obtain its peak value, back bounce data is considered not important and can be not evaluated. The evaluation periods varies for accelerations, forces and Position. Table 10 shows the evaluation periods for the 60 degree lapbelt only configuration. TABLE 10. EVALUATION PERIODS FOR THE FAA HIII 60 O CONFIGURATIONS [3]. CONFIGURATION ACCELEROMETER SIGNALS LOAD CELL SIGNALS POSITION 2 Point Restraint 60 DegreeS 125ms 125ms 125ms In order to maintain the errors unbiased, Olivares [3] compares same ATD responses (curves) in every possible combination. This allows each curve to be the reference data f(t) and also the data to compare g(t). e.g.: whether compare any response from tests -21,-22 and -23 SM compare; 21vs22, 21vs23, 22vs21, 23vs21, 22vs23 and 23vs22 and the maximum errors for shape, delta and magnitude obtained are the measured results for that response (worst cases), which the smaller the value, the better the repeatability of the response. (Appendix A includes the complete error metrics and possible combinations per each configuration) Measured repeatability results of the Lumbar-Pelvis response and seat pan reactions for each 60 degrees configuration are presented in form of bar charts in Figures 30 and 31 and table 11. The bar s height represents the averaged error obtained among all possible errors and the brackets their range from lower limit (minimum error) to upper limit which is the maximum error and measured result. 46

60 FIGURE 30. Repeatability of the lumbar forces and Y moments for all configurations [3]. FIGURE 31. Repeatability of the seat pan forces for all configurations [3]. 47

61 TABLE 11. MEASURED REPEATABILITY OF THE FAA HIII FOR 60 O CONFIGURATIONS [3]. MEASURED REPEATABILITY FOR LUMBAR COMPLIANCE RESPONSES OF THE FAA HIII FOR 60 O CONFIGURATIONS CONFIGURATION TESTS RESPONSES AVERAGE (ERROR) MAX. (ERROR) Measured Result MIN. (ERROR) Part Rigid Seat (14 G's) Part Rigid Seat (19 G's) Part Cushion Seat (14 G's) *Units [lbf, in.lbf] Mag. Shape Delta Mag. Shape Delta Mag. Shape Delta Seat Pan - Fx 8.45% 4.31% % 6.08% % 3.27% 26 Seat Pan - Fz 1.16% 6.26% % 9.64% % 2.11% 11 Seat Pan Res. 0.38% 5.21% % 7.89% % 1.66% 3 Lumbar- Fz 7.12% 5.55% % 7.95% % 1.19% 16 Lumbar - My 18.41% 28.81% % 51.93% % 11.66% 24 Seat Pan - Fx 0.54% 1.20% % 1.31% % 1.11% 3 Seat Pan - Fz 2.61% 2.28% % 3.16% % 1.03% 39 Seat Pan Res. 2.38% 2.08% % 2.87% % 0.95% 38 Lumbar- Fz 3.24% 2.20% % 2.77% % 1.13% 23 Lumbar - My 13.80% 18.39% % 30.49% % 6.82% 11 Seat Pan - Fx 5.57% 5.30% % 8.07% % 1.84% 7 Seat Pan - Fz 6.24% 5.29% % 7.91% % 1.60% 33 Seat Pan Res. 6.11% 5.01% % 7.66% % 2.31% 72 Lumbar- Fz 8.90% 7.03% % 10.99% % 4.98% 130 Lumbar - My 19.02% 21.11% % 35.50% % 3.03% Test Data Analysis and Discussion As is mentioned in previous section 3.2, SAE aircraft committee working group ARP 5765 [3,21] is currently defining, at the moment of writing this thesis research, the minimum acceptable limits of correlation in order to determine the fitness of the ATD responses. As is defined in SAE document repeatability results (errors) above the correlation limits are unacceptable [3,21]. The error metrics gives an error when two responses are compared, where the smaller the error the better the repeatability of the response [3]. Table 12 shows the maximum errors obtained from the error metrics analysis and the minimum correlation limits (bold characters). Observe in the same Table that not all channels tabulated need to correlate with the current limits. Finally note that the sled horizontal acceleration is also presented due the importance and influence of the signal with the ATD responses and their repeatability. 48

62 TABLE 12. FAA HIII RESULTS VS. CORRELATION LIMITS FOR LUMBAR RESPONSES [3,21]. REPEATABILITY OF THE LUMBAR AND SEAT PAN COMPLIANCE RESPONSES FOR THE FAA HYBRID III (MAXIMUM ERRORS OBTAINED) 60 DEG. LAP- BELT-ONLY. PART RIGID S. 14 G S 60 DEG. LAP- BELT-ONLY. PART RIGID S. 19 G S 60 DEG. LAP- BELT-ONLY. PART CUSHION 14 G S Error Limit Error Limit Error Limit Maximum Shape error (%) Maximum Magnitude error (%) Maximum Delta error (G s, lbs, lbs in) N.R.: Not required Sled Ax 4 N.R. 1 N.R. 4 N.R. Lumbar Fz Lumbar My 52 N.R. 30 N.R. 36 N.R. Seat Pan Fx 6 N.R. 1 N.R. 8 N.R. Seat Pan Fz Seat Pan Res 8 N.R. 3 N.R. 8 N.R. Sled Ax 4 N.R. 0 N.R. 2 N.R. Lumbar Fz Lumbar My Seat Pan Fx 13 N.R. 1 N.R. 9 N.R. Seat Pan Fz 0 N.R. 4 N.R. 10 N.R. Seat Pan Res 1 N.R. 4 N.R. 10 N.R. Sled Ax 0.5 N.R. 0.1 N.R. 0.2 N.R. Lumbar Fz N.R N.R N.R. Lumbar My N.R N.R N.R. Seat Pan Fx 84.9 N.R. 7.9 N.R N.R. Seat Pan Fz 9.0 N.R N.R N.R. Seat Pan Res 11.0 N.R N.R N.R. According to the results shown on the table12, all FAA Hybrid III ATD lumbar pelvis responses required to compare are below the minimum correlation limits. Only the Lumbar Fz for Part cushioned (highlighted in red) exceeded in one percent the allowable error (+1%) and is considered irrelevant. Therefore, the FAA Hybrid III 50 th male achieves a high degree of repeatability for the three 60 o configurations based on current correlation limits [3, 21]. The reliability of the FAA HIII ATD is also demonstrated due the stable behavior that the Hybrid III demonstrated when tested under different pulses. Observe in Figure 32 that the FAA 49

63 HIII passes without inconveniences the pass-fail spine compression criterion when tested on a rigid seat with 14 G s (Part transport airplanes). However, when the same dummy is tested for the same pulse severity but with cushion, the ATD fails the spinal compression criteria in all three tests. Likewise for Part (General Aviation) with a 19 G s pulse the dummy fails the mentioned criteria again in all three tests. Figure 32 show the averaged responses for each configuration and the differences among Lumbar vertical forces for each configuration when compared with the spinal compressive Pass/Fail criterion. FIGURE 32. Averaged compliance lumbar responses all Parts [3]. Another important observation is the difference obtained in the lumbar vertical load for part when compared rigid seat tests versus cushion seat tests (see Table 13 below). While the rigid seat tests easily pass the criterion, the use of a non-flotation aircraft cushion gives 50

64 opposite results. It is predictable worst results when a thick-soft element is introduced between the pelvis and seat pan compared with a rigid seat only. The thickness of the cushion and foam properties allow the spine to absorb, thus record more pure axial load until begins rotating. For the low 14 G s pulse the cushion adds to the spine almost 1200 lbs. (averaged load) of undesirable load compared with the rigid seat tests resulting in unacceptable lumbar vertical responses according to the spinal compressive criterion. The 14 G s cushioned results are also larger than a more severe 19 G s rigid seat tests. Despite the non-pass criteria results of the cushioned tests, all three developed similar values and achieved high degree of repeatability. The robustness of the results might be of great use for aircraft seat designers and manufacturers for possible predictions and answer what if questions whether cushion is added. TABLE 13. COMPARISON OF THE FAA HIII WITH CUSHION VS. RIGID SEAT FOR PART TEST #. ATD Serial# ANGLE (deg) LOAD. SEAT TYPE CRASH PULSE LUMBAR LOAD (lbf) AVERAGE PEAK LUMBAR Fz SPINE CRIT. (lbf) (lbf) FAA HYB III g Rigid FAA HYB III g Rigid FAA HYB III g Rigid FAA HYB III g Cushion FAA HYB III g Cushion FAA HYB III g Cushion

65 The last final observation but also relevant is that the FAA Hybrid III maintains a linear trend behavior for the lumbar vertical force among all 60 degree configurations. The reliable spine behavior of the dummy might help to predict possible lumbar vertical results for other and stringent severity pulses at a first phase of a seat design and again being a great of use for seat designers and manufacturers. Next figure shows the three configurations from passed to failed spinal criteria averaged results, each test result, and the linear behavior among them. FIGURE 33. Lumbar vertical forces for all 60 degrees configurations. 52