Survivable Impact Forces on Human Body Constrained by Full Body Harness HSL/2003/09 Prepared by Harry Crawford for the Health and Safety Executive © Crown Copyright 2003Survivable Impact Forces on Human Body Constrained by Full Body Harness Harry Crawford C Eng M I Mech E 29 Tulliallan Place, East Kilbride, Glasgow, G74 2EG This study was commissioned in an effort to reduce the potential for leg and ankle construction and roofing workers employed in the construction and cladding of ‘low roofs" structures of this type the workers often secure their hamess lanyard t0 a strong point, or anchorage, at “foot level, If a fall should occur, the combination of 2m lanyard length plus extension of the energy absorber and the height from harness attachment to the worker's feet ean exceed the height from the structural anchorage to the floor. ‘The worker's feet may strike the ground or floor whilst the energy absorber is still deployi It has been suggested that reduction of the fall-arrest distance may reduce the potential for these injuries, but the laws of physics indicate this cannot be achieved without consequent increase of anrest forces on the body. ‘This study investigates the possibility of raising the level of the fall- arrest force. It also suggests alternative solutions. Analysis of the medical, physiological and other scientific literature regularly shows up the fa that “the learned” talk to “the learned” in terminology foreign to other educated readers. ‘This paper seeks to “demystify” the information and make it available and understandable to those whose interest is industrial fall-safety. The study was funded by the Health & Safety Executive. Its contents, including any opinions and or conclusions, are those of the author and do not necessarily reflect HSE policy.ACKNOWLEDGEMENTS The author wishes to thank his assistant Lorraine for her patience; Ann Simpson, The Knowledge Key Ltd., for her information retrieval skills; Gillian Hution, Dstl Knowledge Services, for her enthusiasm in tracing NATO papers: David Riches, Safety Squared, UK, for his assistance in tracing early papers from the RAF Institute of Aviation Medicine; cutive, for the idea to pursue this study, and for proof David Thomas, Health and Safety reading,1 14 12 13 134 1.32 133 134 135 2 a4 32 33 44 42 43 44 45 46 CONTENTS, ACKNOWLEDGEMENTS EXCUTIVE SUMMARY INTRODUCTION BACKGROUND OBJECTIVES DEFINITIONS ‘Acceleration due to gravity (g) The “G" system of units Jott Fall Factor Physiological Orientation A BRIEF HISTORY OF IMPACT TESTING EVENTS LEADING TO CEN SELECTION OF 6kN MAXIMUM ARREST FORCE INFORMATION GATHERING PRIOR TO EN STANDARDS. U.S.A. AND CANADIAN REQUIREMENTS. IMPLEMENTATION OF THE EN STANDARDS DATA FROM THE PRESENT STUDY LITERATURE SEARCH EFFECT OF JOLT STRENGTH OF VERTEBRAE AND INTERVERTEBRAL Discs POTENTIAL FOR CERVICAL SPINE INJURY POTENTIAL FOR THORACIC SPINE INJURY POTENTIAL FOR LUMBAR SPINE INJURY " 12 13o 52 53 54 55 10. 10.1 CONTENTS OBSERVATIONS ON LANYARD, ENERGY ABSORBER AND HARNESS EXTENSION IN A FALL LANYARDS ENERGY ABSORBERS HARNESS EXTENSION COMBINATION OF FACTORS AFFECTING FALL-ARREST HEIGHT OBSERVATIONS ON FALL-ARREST CALCULATIONS: CONCLUSIONS AND RECOMMENDATIONS: ANNEX ‘A’ - FORCE/TIME TRACES FOR LANYARD/ENERGY ABSORBER ASSEMBLIES. ANNEX 'B' - TABLES OF CALCULATED DROP HEIGHTS FOR VARIOUS LANYARD AND ENERGY ABSORBER COMBINATIONS ANNEX 'C’ - HIGHLIGHTS OF PAPERS ANNEX 'D' - BIBLIOGRAPHY PAPERS CITED BY ABOVE RESEARCHERS BUT NOT STUDIED HEREIN 14 14 14 15 "7 18 19 24 25 39 87 60EXECUTIVE SUMMARY The object of this study was to determine if it is medically supportable to develop energy absorbing devices with arrest force greater than the present CEN standard 6kN maximum advised for wearers of industrial full body hamesses. ‘The study was initiated following reports that workers on low roof’ can be exposed to a peculiar fall hazard. During steelwork and roof laying phases on low-roof constructions, e.g. factories and shops, the linear advancement of the project requires regular relocation of the worker's lanyard anchorage. In work of this type it is common for the worker to anchor to the strueture at feet level. Such @ low anchorage can result in ankle and leg injury if, in a fall, the combination of lanyard energy absorber, harness assembly and body height exceeds the height to the floor or ground, Reduction of fall height entails increase of fall arrest forces. A major feature of the study was the gathering of information on impact tolerance levels on the human body, particularly seat-to-head data, Although there is much aneedotal information from climbing. diving. football and other sports activities, these sources were of litte help due to lack of information on physical forces involved ‘The study therefore concentrates on relevant, scientifically measured data from biomechanical earch by the National Aeronautics and Space Administration (NASA) and the Advisory Group for Aerospace and Development (AGARD) - a group serving the interests of the North Atlantic ‘Treaty Organisation (NATO). — Included in the NASA and AGARD research are medical/physiological data relating to strength of the spinal column, vertebrae and intervertebral discs. ‘These researches were conducted largely in the 1950's and 1960's but several analyses have been produced in the subsequent years; relevant re lighted throughout the study. ‘The literature is not sympathetic to the notion of increasing present levels of arrest force on wearers of fall body hamesses. A deceleration of 12G is considered survivable in a parachute hhamess, ie. a hamess with torso enclosing straps and shoulder straps. For such hamesses the NASA/AGARD researches indicate a 5% injury risk at 12.1G, but the differing posture, physical fitness levels, hamess attachment location, ‘wearer comfort and other factors have influenced the advisability of 6G as a maximum for users of industrial harnesses. ‘The study wily understood mathematics which show that efforts to reduce arrest distance by increasing the arrest force introduce a law of ‘diminishing returns.. It concludes that the present 6kN limit (EN standards) is a wise choice for body weights in the range 80kg to 100kg, But it is recommended herein that 4KN maximum arrest force is more suitable for body weights in the range SOky 10 80kg, and 8KN max would be suitable for body weights in the range 100kg 10 HOkg. Strong recommendations are made that UK and CEN standards bodies should serio pursue this proposal. The initial purpose of the study - safety on low roofs - is somewhat swamped by the biomechanical information but, included in the conclusions, it is suggested that safety on low-roof work may be best improved by the use of the relatively new innovation of lightweight, portable floor mats. ‘Also included are guidance tables and diagrams illustrating the test performance of various systems and comparisons of these with predicted performance when used by a person. This part of the study highlights ‘stretch’ in various components that leads to unexpected inerease in the fall- arrest height and, as such, provides guidance for installers.tn matters biomechanical’, the work highlights the reported ‘greatest risk’ areas of the cervical, thoracie and lumbar vertebrae; it also identifies papers that comment on injury to internal organs at high levels of 'seat-to-head’ deceleration. ‘The reported strengths of vertebrae and intervertebral dises are shown in diagrammatic form and further listed in the annex "Highlights of Papers. The study entailed the scrutiny of 53 relevant scientific and biomechanical research papers, Brief es of these are included as an annex, along with a full bibliography. a1 INTRODUCTION 4.1 BACKGROUND Until the introduction of European Standards (EN standards) in 1993, the accepted UK standard for harnesses and associated equipment was BS 1397. From its inception BS 1397:1947 “Specification for safety belts and harnesses”, and its revisions over the years 1956, 1967 and 1979, advised performance norms for hamesses and associated equipment where the structural anchorage point was always above the user. ‘The worst ease was considered to be when the anchor point was horizontally in line with the attachment point of harness and lanyard (i.e. it was considered that the worker would not fall further than the length of the lanyard ~ 2 metres maximum ~ fall factor 1.0). BS 1397:1979 put a limit of 1OKN on the arrest force, when tested with a 100kg articulated dummy on a fall of 2 metres. During these development years there was awareness in the trades that a worker may not, in certain work tasks, have an anchor point above or adjacent to the hamess/lanyard attachment. A fall could be from above the anchor point (e.g. a steel erector whose only strong point may, at times, be the beam beneath his/her feet). In such a ease, the fall could be twice the length of the 2m long lanyard, ic. 2 x 2m [fall factor 2.0 (FF 2.0)] = 4m. ‘A working group was set up by CEN (European Committee for Standardisation) in the late 1980's to rationalise the various national standards and determine fundamental norms for ‘all safety’ Personal protective equipment (ppe). The group investigated several issues, including the choice i test dummies (sometimes test-mass) and the consequences of 4m falls on industrial workers, The working group recommended to CEN and the EN standards committee (CEN/TC160) that a {all-arrest limit of 6KN be set for EN standards. EN 355:1992 (BS EN 355:1993) “Personal protective equipment against falls from a height ~ Energy absorbers” was therefore introduced, with provision for arresting a fall of 4m with a maximum arrest force of 6KN.* To accommodate the necessary extension the assembly of energy absorber and associated lanyard was permitted to extend by a maximum of 1.75m when tested with a 100kg mass on a free-fall of 4m. Designers and manufacturers of these kinds of ppe have addressed the norms by providing energy absorbers, usually of tear-stitch’ or *tear-ply” construction, operating at around 4KN arrest force. ‘At such arrest force, the energy absorbers tend to take up much of the permitted 1.75m extension. Energy absorbers designed for operation nearer the 6KN limit would be expected to require less extension in a fall-arrest event. Energy absorbers designed to operate above GkN would require yet less extension, hence their viability is an object ofthis study. * The EN standards committee CEN/TC160 fas applied the same fall-arrest force limit of ‘6kN for EN 353-1 and EN 353-2 arresters on rigid and flexible anchorage lines. The 6kN Timit has also been applied on EN 360 retractable type fall-arresters. 1.2. OBJECTIVES The spe of survivable impact forces on the human body cor ion for the project was “to carry out a literature search, critical review and analysis rained by full harness”. The work was to inchide an “international fiterature search t© cover relevant medical and physiological data and 0 ian fall-related industrial accidents, military jection seat and parachute research (here available), environmental medicine, aviation medicn space medicine, climbing and saifing accidents (and any other sources which may become evident in the course of the search)”. The object of the critical review and analysis w (and jolts) as these relate to injury and/or surviv to accumulate information on fall arrest forces prospects for wearers of full body harness inaccidental fall situations. Where such information or me lacking, the review will address and suggest areas for further res Wiphystological information is Simply put, the work was intended to explore relevant international research and medical/physiological opinion and determine if these would suggest extension of the present 6kN ‘maximum arrest force for a human subject wearing a full-body harness on a free fall of up to 4, 1.3. DEFINITIONS For the purposes of this report the following definit measurement: ions apply, together with SI units of 1.3.1 Acceleration due to gravity (g) Natural act planet, but cleration in a fall due to gravity. This varies slightly according to location on the 5 internationally accepted for calculation purposes as 9.81m/s" 4.3.2 The“ system of units is a dimensional representation of the magnitude of acceleration, expressed as the ratio of the magnitude of measured acceleration to the magnitude of “natural” acceleration due to gravity. (A body experiences 1G [in layman’s terms body weight] whilst standing immobile on the ground. This differs from the “g” system where the immobile body is described as experiencing zero “p"). 1.3.3. Jolt Jolt describes the rate of change, or rate of onset, of acceleration. ‘The term is used to describe how rapidly the peak acceleration is reached. It is often expressed in the relevant biomechanical literature as Gis (G~ Rise Time), sometimes as mis’ 1.3.4 Fall Factor (FF) ‘This is a ratio. Itis the height of'a potential fall divided by th ial length of the lanyard, 4.3.5. Physi logical Orientation The system employed in this paper is the universal physiological standard system recommended by the Biodynamies Committee, AGARD Aerospace Medical Panel, (see Souder, “lmpact"[40)) Upward accel. (positive GG) -+G,_—_Eyeballs down Downward (negative G) -G, Eyeballs up Backward accel 4G. Eyeballs out Forwant accel Ge Eyeballs in Acceleration to right 4G, Eyeballs teh Acceleration to left Gy Eyeballs right Figure 1. Diagram and explanation of physiological terminology in this paper2 ABRIEF HISTORY OF IMPACT TESTING The major thrust in the development of impact testing came with World War I and the ion of high-performance aircraft. Eject ‘were first installed, early in the war, in German fighter aircraft when it was found that pilots could not extract themselves manually in an emergency. Il is reported that these seats were used operationally sixty times. During this time compression tests were carried out by Seigitied Ruff on cadaver vertebrae, to determine the strength of the spinal column under high positive G accelerations over short time-spans (He [22], Higgins [23)). Note: References in italies and/or square brackets [-] /-] apply to summarised in Annex C “Highlights of Papers”. found at Annex D, uthorites and papers The bibliography for this paper is to be By 1943, the Swedish air force had installed an ejection seat in their J-21 fighter. This was an aireraft with a pusher propeller and a high horizontal tailplane mounted on twin booms. With such configuration an ejection seat was @ necessity. Investigations of the strength of the spinal column were carried out by Olof Perey. British interest in impact testing began in 1944, with studies by the Martin-Baker Aireraft Company. ‘These were carried out with tests on an early ejeetion-seat test tower using dummies and human subjects, This work, at an early stage, indicated the risks of vertebral injury. Catapults were designed to lessen the risk of injury, and drogue parachutes were added to stabilise and separate the seat from the pilot, The seat was fitted with a protective visor that was pulled down by the airman at ejection and caused him to attain good spinal posture at the instant of ejection. ‘The Martin-Baker seat was introduced into service in 1946 and its many derivatives have seen service world-wide, ‘The US Air Force started work on ejection seats in 1945 (Henzel, Higgins). The US Navy also found it necessary to develop ejection seats for their aircraft, particularly for incidents on take-off and low-level approach to aircraft carriers. To further this work the navy developed a Rocket Assisted Propulsion Ejection Catapult (RAPEC). Researchers at Wayne State University conducted investigations in the biomechanical properties of the vertebral column, whilst teams at Massachusetts General Hospital and Massachusetts Institute of Technology worked on the properties of intervertebral dises, Several centrifuges and rocket driven sleds were built in the USA for study of tolerance and survival values for “magnitude, duration and rate of change of negative acceleration” (i.e. deceleration). Colonel J.P. Stapp is internationally known for the considerable work he carried out in biodynamies using such sleds during the 1950s. Development of parachutes and harnesses was essential to all of the above progress in ejection seat technology. Opening of drogue and the main parachute canopy on separation of an airman from an ejection seat can cause very high deceleration forces when ejecting from a high performance airerall. Stapp and later researchers advised 12G as an upper limit in a parachute harness [2 {13}. (211, 27), [40] and [43] 16), In time, much of the above research data found its way into the public domain via NASA and NATO scientific papers3 EVENTS LEADING TO CEN SELECTION OF 6kN MAXIMUM ARREST FORCE 3.1 INFORMATION GATHERING PRIOR TO EN STANDARDS, As described in the introduction, 1947 saw the beginning (in the UK) of performance testing of industrial “safety belts” and harnesses to BS1397. Until 1979 these tests were mainly “strength tests”, static and dynamic, to ensure the structural integrity of hamesses and thus to protect the ‘wearer from falling to the ground or floor. ‘There was no requirement to test for arrest forees. By the 1970s there was a beginning of aecess to the information learned from military and aerospace studies. In the UK, the source of such information was the RAF Institute of Aviation Medicine, Famborough (Becton, Ernsting, Glaister and Reader). ‘The 1979 version of BS 1397 reflected this growing access in its test specification of 10G maximum for pole belts, 5G maximum for general purpose safety belts and chest harnesses, and 10G for general purpose salety hamesses. A special provision of 12.5G maximum, with full body hamess, was made for coal- mine riggers who had a preference for chain lanyards. All of these 1979 version tests were carried ‘out with an “articulated anthropometric dummy” of 100kg. It és noteworthy that only full body ‘harnesses are now considered suitable in fall-risk situations. With the Treaty of Rome, 1975, and establishment of the European Community (later European Union) and the requirement for removal of barriers to trade, eame the setting up of CEN. During the 1980's, working groups were formed with the aim of rationalising the various national standards, “Their object was to facilitate the work of the later CEN technical committees and the formulation of unified standards to satisfy the European Directives. At this time the work of Stapp [43], Eiband [14] and others was becoming more accessible through NASA and NATO sources. ‘Some of this early work was available to the CEN working group concerned with fall protection. 200 : ‘Area of severe 4 so ainey 40 ‘Area of moderate ey weet Ejection seat Limits Voluntary kumar uninjured DECELERATION g units r 00 oh OlSCSSCOS Fay DURATION secends Figure 2. Survivable abrupt positive G (+G,) impact, from Eiband 4Figure 2 is a presentation of survivable abrupt positive acceleration ¥s. duration, based on data of Biband, The figure uses log-log ordinate (vertical) and abscissa (horizontal) scales. Shown are the zones of voluntary human (uninjured) exposures for subjects secured with lap and shoulder strap jection seat design limits, and zones of probable moderate and severe injury based on research with humans, hogs and chimpanzees. Measurement of uniform deceleration of the drop test vehicle was taken at the seat level (not on the subjects). The fiterature acknowledged that the uninjured, undebilitated, voluntary human exposures shown were with physically fit humans who were secured at the Seat and shoulders. The available medical and physiological information related almost exclusively to studies with military personnel. Other data and medical/physiological opinion indicated that |2kN was the “maximum of tolerance” for fit men at parachute canopy opening (Amphoux [2], Beton (6, Delahaye {13}, Hearon & Brinkley [21]. Kazarian [27], Snyder [40], Stapp [43)). ‘The working group (convened by Dr Maurice Amphoux) took account of the following factors: © Military parachute hamesses are designed with greater torso constraint than industrial harnesses, ic. there is a greater risk of upper spine (cerebral vertebrae) injury in industrial hamesses. due to Nexon. © Industrial workers, in most eases, do not have the high level of phy military personnel, al fitness required of ‘+ Industrial workers include a probably wider age-band than military personnel exposed to fall arrest risks. The literature indicated deterioration of the spine for most at ages beyond 40 years (White & Panjabi [52], later Yoganandan [33)). + The proposed increase in overall fal-height from 2m to 4m would cause increase in the duration of exposure to fall-arrest forces (it should be noted that Franee. through AFNOR standards, had adopted 6G maximum for FF 2 with 2m long lanyards during the 1980's). Such inerease in the duration of exposure to impact would project fall arrest results beyond the forve/duration data of Eiband and Stapp and their definition of impact (those researchers considered “impact” to be an event that did not exceed 0.23) ‘These factors were accepted by the technical committee CEN/TC160, and 6kN was adopted as the maximum arrest force for fall-protection devices used with industrial full-body hamesses. The same norms have since been adopted for the relevant ISO standards. 3.2 U.S.A AND CANADIAN REQUIREMENTS Ameri Standard 2359 [4] and the Canadian Ontario Ministry of Labour requirement for fall protection all provide for 6ft, ‘or 1.0 falls. All three permit a maximum arrest force Of SKN. The OSHA-required dynamic performance test fora lanyard with energy absorber employs a mass of 100kg. ‘The free fall hei ed as “6ft 1.8m)" and the maximum permitted extension is "3.51 (.07my The 2359 requirements are very similar. Test Mass 100kg, free Fall height “6M (1.829m)" and the maximum permitted extension "42 inches (1.067m)”.3.3 IMPLEMENTATION OF THE EN STANDARDS Since the adoption of the EN standards, in 1993, it has been observed that the preference for arrest force - whether with energy absorbing lanyards, inertia reel (self-retracting) devices or guided type artesters has been somewhat below the permitted 6KN level. Most UK-manufactured energy absorbers operate at around 4KN 10 4.5KN. as can be seen at Annex 'A' “Force/Time traces for lanyardenergy absorber assemblies”. Inertia reel devices and guided type arresters tend to operate at around 4.5KN. © 3 with the ian norms, These collected data on popular UK energy-absorbing lanyards are compared in figu CEN and ISO performance norms. Also shown in the figure are the USA and Cans and the plot of the Eiband/Stapp human voluntary exposures, Range ef ejection seats Eiband/Stapp human voluntary exposures (uninjured) ates ees ieee ramet ee eayre teat USA and CANADIAN pevmitted on HBS m —~ fal @ FF VO with 100kg mass CEN #150 permitted on 4m fall | PARR SORT OFF 20 with 100 kg mass SFr we al Sooo ooo TIATAITI PITTI POPPA OR, popular energy absorbers | on 4m fal@FF 2-0 with | 100 ky maze L ° L 1 1 1 L on o2 08 Om os oe Time - seconds Figure 3. Comparison of fall-arrest force norms with Eiband/Stapp data 64 DATA FROM THE PRESENT STUDY 44 LITERATURE SEARCH The literature search was carried out in conjunction with Knowledge Key Lid, ‘The databases used NTIS: NATIONAL TECHNICAL INFORMATION SERVICE. Consists of summaries of US, government-sponsored research, development and engineering from NASA, DOD, DOE, HUD, DOP, Dept Commerce and some 240 other agencies. FL COMPENDEX: Electronic version of the Engineering Index, which provides abstracted information from the world’s. significant engineering and technological literature. The Compendex database provides world-wide coverage of approximately 4,500 journals and selected government reports and books. SCISEARCH: Cited Reference Science Database. An international, multidisciplinary index to the Institute for Scientific Information. SciSearch contains all of the records published in the Science Citation Index plus additional records from the Current Contents publications. TRI TRANSPORTATION RESEARCH INFORMATION SERVICE. Provides international coverage of ongoing research projects, published journal articles, government reports, conference proceedings and technical papers AEROSPACE DATABASE: Provides references, abstracts, technical documents, books, conferences and reports covering aerospace research and development in over 40 countries including Japan and Easter European nations. It also contains reports issued by NASA, other US government agencies, international institutions, universities and private firms. FEDERAL RESEARCH IN PROGRESS: Provitles access to information about ongoing US federally funded research projects in the fields of physical sciences, engineering and life seiences, PEDS: DEFENCE PROGRAM SUMMARIES. Collection of justification documents that correlate to each defence program element. Literally meaning Program Element Descriptive Summaries, PEDS provide detailed descriptions of specific programs and their costs. MCGRAW-HILL COMPANIES PUBLICATIONS ONLINE: — Provides text for major McGraw-Hill publications covering not only general business but also specific industies, ie aerospace, chemical processing, electronies and construction. ROUP AEROSPACE/DEFENCE MARKETS AND TECHNOLOGY: | Provides articles and abstracts covering all aspects of the world-wide aerospace industry. OCCUPATIONAL SAFETY AND HEALTH: Includes citations to more than 400 journal titles, as well as over 70,000 technical reports covering all aspects of health and safety. SPORTDISCUS: ‘The Sport Information Resource Centre, the database provider, is the largest resource centre in the world collecting and disseminating information in the area of sport and sports medicine, Mi re LINE: Is one of the major sources for biomedical literature with approxima ords added to the database per year ly 400,000 INSPEC: Corresponds to the three Science Abstracts print publications: Physies Abstracts, Electrical and Electronics Abstracts, and Computer and Control Abstracts.AGARD PAPERS: A further source of data proved to be Dstl Knowledge Services, Glasgow. ‘This is a government office and was the source for all of the NATO AGARD scientific papers quoted and cited herein. 4.2 EFFECT OF JOLT Jolt is defined as the rate of onset of acceleration. For the range of acceleration or decelera considered in this study itis probably best explained as ion Initial peak acceler Rise time ion which can be expressed in m/s’ or, as in most of the papers reviewed, Gis. Annex A shows several forcertime traces for popular UK-manufactured lanyardenergy absorber assemblies. The jolt is seen to be the slope of the trace as the deceleration force rises on engagement of the energy absorber. ‘The steeper the slope, the more serious the effect on the body, The human body can tolerate very high jolt levels over very short periods of time, usually because the amplitude, or distance travelled, is small. Jolt becomes a serious problem when the duration oF amplitude increases. For subjects exposed to positive G the means of constraint of the body is important. ‘The spinal column is limited in strength, as many of the references show (Amphoux [2]. Burton et al [7], Delahaye [12,13], Eiband [14], Higgins [23], Jones [25]. Kazarian et al [27], Shaw [39]. ‘Snyder [42]. Stapp [43]. Swearingen et al [47], and Teyssandier [48]). Ve the torso is positively secured, as for example in a modern ejection seat, the basic vertebral strength can be enhanced because the bulk of muscle and general body tissue provides support for very short periods of time (Stapp). ‘The Eiband/Stapp data indicate ejection seat operation in the range I8G 10 22G. Delahaye cites “G forces greater than +15G, for duration of 0.2 t0 0.5 seconds”. ‘On separation from the ejection seat, several researchers (Amphoux, Beeton et al {6], Delahaye, and Kazarian et al) are of the strong opinion that 12G should be considered the maximum deceleration in a “parachute harness”. It is noteworthy that US military specification 9479 ‘considers 12.1G to be the 5% “probability of injury level "(Kazarian et al). {t appears that most of the workers in the area of 4G, research have concentrated on tolerance of the spinal column, Where the subject individual is less securely constrained (as in an industeial hhamess) the risk of spinal damage increases with high G forces ~ and with high jolt levels! There is greater risk of flexion ofthe spine, and consequent injury. There is also a risk of internal injury due to inertia of the major organs. Figure 4 illustrates how the heart and lungs are separated from the lower organs (liver, kidneys, intestines etc) by the diaphragm. In a fall-arrest event a reasonably erect spine is, to an extent, supported by the bulk of the musculo-skeletal system, but the inner organs are more “loosely” suspended and thus more affected by jolt, Swearingen [47] tells of crashes where helicopter pilots died of ruptured aortas and argues that 10G measured at shoulder level is a survivable maximum for helicopter pilots. Wallace and Swearingen [51] also describe a tragic case where 6 young men died in a light aircraft that hooked a power line and came to earth “ita flat attitude”, "None of the victims showed any external sign of injury. All had died from “severe impact trauma to intemal organs (brain, heart, liver, spleen ete)”. In both reports the levels of positive G and jolt were unknown, but the nature fof the injuries alerts the reader to the vulnerability of the inner organs in conditions of high positive Gand jot.~~alodominal acvta, Figure 4. Major organs above and below diaphragm 4.3 STRENGTH OF VERTEBRAE AND INTERVERTEBRAL DISCS | cervical thoracic ai rs Lumba sacrum cotcgx Figure 6. The Spine Figure S is drawn from medical literature, It illustrates the major zones of the spinal column, ‘The upper 7 vertebrae form the cervical spine, the next 12 the thoracic (sometimes called dorsal), and the lower 5 the lumbar, ‘The sacrum and coccyx. complete the structure. Intervertebral discs. separate the vertebrae, starting at approximately 2 to 4mm thick in the cervical region and increasing, to approximately 12mm thick in the lumbar region. ‘The average length of the spinal column is of the order of 750mm, 30% to 35% composed of dises. The strength of vertebrae increases from top 10 bottom of the spine, remarkably in proportion 1 the weight of the trunk above any given vertebra, Henzel [22] quotes Ruff and Stech findings that the resistance to positive G also rises consistently. throughout the column, except for a sinall “dip* in the region T12 to 1.2 (see also Figure 6). The literature (see 4.4, 4.5 and 4,6) indicate C5-C7 to be most vulnerable in the upper spine, and T12- LI to be most often damaged in the vulnerable T8- L3 lower spine range.Figure 6 is a presentation of information on vertebral strength drawn from the literature studied. The C3 to LS data are drawn from White and Panjabi [52], who have compared the C3 - T12 work of Messerer (1880), LI - LS work of Perry in other papers Perey] (1957), and L-4 work of Bell et al (1967). Higgins’ [23] data from Evans et al (12 — LS) and Massachusetts General Hospita/MIT (2 — LS) are seen to be lower than White & Panjabi. Sances et al [36] and Teyssandier [48] are seen to arrive at higher results, Henzel [22] and Higgins separately quote data (T8—L4) from Ruf?’s early work. Ruf?s results from the 1940s are cleaely higher than all of the other researchers. Rulf states his work to be based on body weight of 75kg. Perey identifies age difference. White and Panjabi explain the difference in results as probable differences in the design of the tests, and age and condition of cadavers. The other researchers do not appear to record age or body size. Henzel’s paper is most illuminating. He compares the work of Ruff and another researcher, Stech, who noted that other serious damage occurs in vertebrae before they reach “breaking” point Stech had observed that vertebrae, under compression, pass through “end plate fracture”, then the “limit of proportionality”, next “yield point” and finally “breaking point”. ‘These features are in some ways similar to our understanding of the strength of engineering materials. End plate damage in a vertebra is caused by high dise pressures (using the analogy of a ‘metal can, the lid and base represent the end plates) Results on intervertebral dises have been added to illustrate the considerable strength of dises in the lower thoracie and lumbar regions. It ean be seen that the dises are similar in strength to the vertebrae in the cervical and upper thoracie region. Dise strength increases progressively down the spinal column until, in the lumbar region, the discs are approximately 3 times stronger than the vertebrae (Burton et al [7] and Sances et al [36]).. This information helps the reader see that, under impact, vertebra damage (e.g, end plate damage) can occur before dise damage (see Figure 6, Sances “DISCS’). TTT Tt rT =a t on S + | ~ --_ SANCES | +t, = R(bises) S Se Pp IN S fit = 1 a SPS | ry =] + a t : + - sre piece PANTABL + (RUFF) 4 uicews. 8 (evans) A 7 x £ LY SSANDIER: Seni) DECOULX, RIEUNAL) bof op op yg 2 7 Ne 16 2 is Figure 6. Strength of vertebrae and intervertebral discs loMost of the above fiterature is academic in nature and has been of considerable guidance in the Mn of air force task tolerance levels, i.e. for physically fit personnel, Henzel argues that the setting of tolerance limits for airerafl ejection seats and subsequent parachute forees accepts that there isa ‘military risk level’, i.e. a risk of injury. There is no suggestion of tolerance levels for the wide variation of users of industrial hamesses. 4.4 POTENTIAL FOR CERVICAL SPINE INJURY ‘The cervieal spine, oF neck area, consists of 7 vertebrae and is the most manoeuvrable region of the spinal column. In flexion, tateral bending and torsion it is by far the most flexible. Readers who wish to pursue an academic study of the biomechanics of the cervical spine are referred to the work of Ashton-Mitler & Schultz [5] and White & Panjabi [52]. Information on cervieal spine injury due to impact is almost entirely drawn from military and aerospace studies. NASA has conducted considerable work in this field and much of this is to be found, in addition to European research, in papers by NATO group AGARD [Advisory Group for ‘Aerospace Research & Development] (AGARD paper [1] , Burton et al {7}, Delahaye {13}. Ewing and Thomas {16}. Snyder [40], States (44) The cervical spine is capable of sustaining high forees when the head and vertebrae are in optimum alignment (Burton et al, Chub {8}. Schall {37}, Yoganandan (53}). Schall deseribes flying manoeuvres where airerew with helmets can be exposed to 65kzf (636N) force on the cervical vertebrae, He also notes the ability of some native Africans to cary “91k without difficulty”. When optimum alignment of the head and neck is not achieved, the potential for injury is considerable. Burton and his colleagues in NATO (RTO-TR-4) (see Burton et al) describe cervieal spine injuries in manoeuvres of positive 4G. ‘These usually occurred when the airmen were making observation head-movements. Deakin [11] reports similar injuries on 8G tums. It is of interest that most of the airerew thus injured were able to return to duty after medical treatment, some to lighter duties. Burton tells of the advantages of flyers with muscular necks and, at chapter 14.7, describes effort by several air forces to introduce physical activities to strengthen neck muscles. He concludes. however, that such “an exercise regimen” has not yet been developed that will “produce balanced neck support”. Henzel [22] refers to work published by Jefferson in 1928, long before the advent of high performance aircraft (or of full-body harnesses). Jefferson had studied 2006 eases of fall-related spinal fracture. Of these, 28% involved fracture of the C5-C7 vertebrae, the lower cervical spine area (see Figure 5). This information is rather historical but, considered together with the above data, it demonstrates the vulnerability of this region of the spine to flexion. Many of the more recent researchers express concern about cases of whiplash or flexion of the upper spine, ustially ‘caused when the flyer's head and neck are not in optimum alignment at the moment of impact or high G manoeuvre, Delahaye goes further by describing hyperextension of the upper spine. sulfered at times by parachutists. This is the name given to the mechanism where the inertia of the head causes the chin to strike the sternum (frontal bone of the rib cage), Followed by “nodding”, a to-and-fro action of the head and neck, the “classic bell-ringing” motion. Teyssandier [49] makes the observation that cervical spine injuries among French paratroopers and sports parachutists have gradually reduced in number over the years. He associates this. with developments in parachute design and folding techniques that have led to reduced “shod! opening, His view is that the reduetion in injuries is in ditect ratio to the reduced “shoek’ Delahaye and Mi © [13] AGARD-AG-250(Eng) Chapter 5, discuss both parachute “opening shock” and “landing shock”. They observe that cervical spine injuries due to opening shock have been reduced with the introduction of “eanopy first opening” instead of “canopy first packing”In the light of these accounts, there would seem to be some concern that EN 361 “full body harnesses” permits a post-drop angle of 50” between the “longitudinal axis of the dorsal plane of the torso dummy and the vertical”. International Standard ISO. 1033-1 “full-body hamesse permits a post-fall angle of 45° for some classes on a similar drop test. ‘There are no data on. cervical spine injury on “survived” falls with hamesses to those speci ich large 's from the vertical would seem to present possible whipl dangers. Most hamess manufacturers advise adjustinent of the harness so that the attachment point is high fon the back, between the shoulder blades. It would, however, appear prudent to have a comprehensive investigation of the relationship between harness (plural) drop test results. with torso dummies and the suspension angle for a wide range of human body size and shape. ISO 1033-1 already includes a suspension angle test with “a minimum of three people”. All of the evidence points to there being a lower spinal injury rate when the spine is near vertical at the ‘moment of impact ina hamess. ‘The ISO specification suspension procedures would appear to offer “a starting point” 4.5 POTENTIAL FOR THORACIC SPINE INJURY The thoracic spine (T1-T12 in Figure 5), particularly TI to T8, has a good range in torsion and Alexion, but the lowest lateral bending range of the spinal column (Ashron-Miller & Schultz (5) Again, the data are virually all from NASA and NATO military or aerospace research. Most injuries to the thoracic spine are recorded in the lower thoracic area, T8-TI2, and are usually associated with injury to the lumbar L1 and L2 (Delahaye ef al [13], Henzel [22), Jones etal [25] and Snyder [41]). Jones et al dealt with US Navy ejection injuries 1958-63 and UK aircrew ejection injuries 1949-60, all with Martin Baker seats. ‘The ejection accelerations ranged from 17G to 22G. The paper indicates that most of the uninjured ejections were at the lower end of the +G, range, and the high proportion of injuries were at the upper end of the range. The greater proportion of injuries for US and UK airerew were in the T8-L.1 region, several flyers suffering 2 fractured vertebrae. The same study recorded Swedish injuries 1957-60 with the SAAB seat Injuries with this seat were mainly in the T4-17 region, with several flyers suffering 4 and 5 fractured vertebrae, ‘There is no information on the acceleration levels for the SAAB seat, but itis to be expected that it lay within the envelope indicated at figures 2 and 3. As with the cervical spine study, Henzel refers to Jefferson's (1928) work on the frequency of T12 and LI fractures. Of 2006 cases, 191 (9.5%) were TI2 fractures and 194 (9.79%) were LI fractures. The literature recommends a maximum of 12G in a parachute harness at canopy opening (Amphoe [2], Burton et al [6]. Delahaye & Metges [13], Hearon & Brinkley [21), Karazian 127), Snvder [40] and Stapp [45]). As explained at 3.1 above, CEN accepted 6kN (i.e. 6G on a 1OOkg person) as a maximum for industrial workers, fo take account of the wide ranges of aye and physical fitness. The 6kN level of maximum arrest force, for 100kg body weight, also tak account of the probable uncontrolled nature of industrial falls and the likelihood that the spine of | the subject will not be in ‘optimum alignment’. But, as 5.2 herein shows, 6kN force would not appear suitable for workers whose body weight is considerably below 100kg.4.6 POTENTIAL FOR LUMBAR SPINE INJURY The lumbar spine (L.1-LS in Figure 5) has good flexion capacity, moderate lateral bending abili but very low torsion capacity (Ashfon-Miller & Schultz). Both Ashton-Miller & Schultz [5] and White & Panjabi [52] describe the considerable wall of muscle-tissue surrounding, the lumbar region, ‘The: ely involved in maintaining upright stance and siting stability. White & Panjabi also point out “they also contribute to the very high loads to which the lumbar spine is subjected”. The centre of gravity of the trunk is in front of the spinal column so that considerable muscular effort is required to keep the spine in ‘optimum alignment’. ‘The muscular effort thus induces dise forces that are much higher than would be expected from the weight of trunk above the dise (see Menzel (22J and Sances et af [36)). Once more, the data on lumbar spine injury and pain are virtually all from NASA. and NATO military or aerospace research, The literature evidence is abundant that the intervertebral dises in this region of the spine are 3 times stronger than the vertebrae (Burton et al [7], Sances et al) This probably accounts for the number of researchers whose interests concentrate on the vertebral strength of the lower spine (see Figure 6). The work of all of the researchers studied (including Henzel, Higgins, Sances, Teyssandier, White & Panjabi) would seem to indicate that the spine, including the lumbar region, should not be capable of surviving ejection seat forces. Even the selection of 12.1G as the US military 5% “risk of injury” level in a parachute harness goes beyond the measured strength of the vertebrae. The accepted reason for the apparent strength of the lumbar spine at these acceferations is the support given by the surrounding muscular wall and ‘other body tissue. Shaw (39] mentions fractures of lumbar vertebrae on 25G seat-acceleration in early German work Teyssandier [48] has conducted a survey of 1.468.399 jumps by French paratroopers. Of these, there were 219 who suffered fractures of the spine ~ 76% of them occurred in the lower back between T12 and L3, ‘The reported French parachuting accidents were mainly due t0 “bad landings’. Delahaye & Metges [13] discuss the effects of wind speed on parachute landing velocity. They point out that “landing shock” varies “with the square of the horizontal wind speed” and that high wind speeds account for a high proportion of landing injuries. In addition to ankle and leg injury, the spinal region at greatest risk in these circumstances is T12 to L3.5 OBSERVATIONS ON LANYARD, ENERGY ABSORBER AND HARNESS EXTENSION IN A FALL 5.1 LANYARDS Lanyards are generally manufactured from polyamide or polyester webbing or rope. ‘Those made from webbing usually have very little stretch in the fall-arrest force range. Ropes are usually of [2mm or 6mm diameter and may have considerable stretch in the fallarrest force range. Generally speaking, rope lanyards give rise to a “gentler” increase in force on arrest. The ‘stretch’ time to reach peak deceleration is generally longer than for webbing, ic. the rate of onset (or ‘jolt”) is lower. Where the risk analysis for a task shows there to be no danger of striking structure below the worker, the use of a rope lanyard (with energy absorber) may be preferred. If there is a need to keep ‘stretch’ to a minimum a webbing lanyard is probably to be preferred. 5.2 ENERGY ABSORBERS Energy absorbers tend to be mainly of tear-ply” or “tear stiteh’ design. Several manufacturers have produced designs that, on dynamic testing, have a reliable Mla” peak arrest force. Test traces for three popular, different, designs are shown at Annex A. ‘The energy absorbers and their associated webbing lanyards were dynamically tested in accordance with EN 355 (4m drop with a 100ky mass). ‘The foree/time traces have an apparently rapid rate of onset of acceleration (jolt), usual with webbing lanyards. ‘The results were: No.1 Peak force 4.6KN, flat 4kN arrest force, jolt 265G/s (2602mvs"), No.2 Peak force 4.62kN, reasonably flat 4kN arrest force, jolt 288G/s (2824m/s"), No.3 Peak force 4.8KN, slightly erratie arrest force between 3.5KN and 4.SKN, and jolt T18GIs (1153nVs It can be seen that the designers were “playing safe” within the 6KN limit. Nos. | and 2 had more rapid rates of acceleration onset (jolt) than No.3 (i.e. their “slopes” are steeper), Energy absorbers in Europe should comply with EN 355, which requires capability for arrest of a 100ke drop mas fall of dm. Anallowance of 1.75m is made for extension of the enerey absorber in the arrest phase, thus the absorption capacity must be 5.75kI minimum. The ISO 1033-2, Type 2. requirement is identical, whilst the Type | (max. free-fall 1.8m) has an arrest- force limit of 4KN and maximum extension of 1.2m (3kJ energy absorption capacity minimum). ona fret The OSHA [30] and ANSI 2359 [4] requirements for full hamess are specified as SKN maximum With a 100kg mass free-fall of 1.8m, ‘The maximum permitted extension is 1.7m. The author has long been perplexed that the industry supplies only one solution for a mul variable problem. ‘The standards all employ a 100kg mass on dynamic testing. This does not take account of the huge range of body weight of workers in the field, nominally SOkg to L40kg (7.8st to 22st), possibly greater. ‘The positive G effect with a 6KN energy absorber is compared with 4kN and 8kN absorbers in Table |:Table 1 Deceleration levels on various body weights at 6kN, 4kN and 8kN arrest force Body mass (hg) | 6kN arrest force (6) | akN arrest force (G) | SAN arrest force (G) 50 122 #1 6s oo 10. 68 13.6 70 37 Be 116 80. 7.6 51 10.2) | 0 os a5 amo 100 6 al - 0 [36 3.7 7 120 Su 34 — ins 47 zi 1 a 39 Table | indicates that an energy absorber operating at 6kN maximum is probably best suited for workers of body weight 80kg to 100kg, in fall exposures of 4m (FF 2.0). In the range of body weight 50kg to 80kg the worker would be advised to seek an energy absorber of 4KN maximum, In the range of body weight 100kg to 140kg the worker would be advised to seek an energy absorber of SKN maxi h a range of energy absorbers is not available to the industry a this time. (There have been ‘rumblings’ from the field that large people. in particular, are not well served by the present standards for energy absorbers or other fall-arresting devices designed for “standard” dummies of 100kg). ‘The author strongly recommends that attention be given to bod! weights outwith the 80kg to 100kg person. Designers should ensure that devices are capable of absorbing the potential energy of the heaviest recommended user in each range. ‘There is yet another factor to be considered. Few manufacturers supply information on actual test results. “They may argue that there are commercial reasons for not stating the actual test performance of given energy absorbers (Ie. actual arrest foree and extension), but safety engineers responsible for risk assessments would be helped considerably if this information was routinely made available. 5.3 HARNESS EXTENSION The progression from knotted rope to waist belt to full-body hares has been a long. one. Pressures of safety directives and law. along with the retirement of a reluctant older generation. have brought on the acceptance of full-body hamesses. Although designed primarily for safety, hhamesses are intended to be comfortable to Wear. Most are provided with ample adjustment. The growing trend in industry is t0 issue each worker with a *personal” hamess and other safety kit Safety directives require that manufacturers provide Tull instructions on proper adjustment of @ harness, but a visit to most construction sites will give the visitor an idea of how seldom the fitting, instructions are heeded. The author has seen many workers with loosely fitted hamess (“for comfort”), and with the lanyard attachment “D” ring somewhere in the middle of the back, instead of between the shoulder blades per the fitting instructions, For the best possible outcome in the event of a fall, the preferred position for attachment of lanyard to hamess is atop the shoulders, as with a parachute harness. Analysis of the literature indicates that the ‘upright’ arrest and suspension angle will make for optimum spinal alignment, and least risk of cervical spine damage (due to whiplash or “bell-ringing’). Wit this configuration the many researchers covered in the literature search agree that 12G is a “reasonable risk” maximum (the US military 5% risk of injury limit is. 12.1G}. tt must be borne in mind that such 12G limit is for physically fit, young, military personnelBut atop-the-shoulders attachment is not practical for users of industrial harness. It is inconvenient in most working environments and there is a general fear of injury to the head and ears in “awkward” falls. For this reason the high dorsal (between shoulder blades) attachment position is to be preferred. In addition to being least inconvenient in carrying out a task, this, attachment (on a well-designed hamess that is correctly fitted) will provide reasonable spinal alignment in a fall, ‘The same well-designed, correctly-fitted, hamess will provide a satisfactory post-suspension position whilst attempting self-rescue or awaiting rescuers (but see Orzech [29] and Seddon [38] on risks when person is immobile due to injury or unconsciousness). When adjusted correctly a hamess must, of necessity, have “slack” for body movement. Such slack’ will usually add to the overall height of a fall (see Figure 7) when the harness straps tighten inder load. This additional height from feet (© hamess attachment point is often missed in estimates of clearance beneath the worker. Conversely, an incorrectly’ fitted harness and large suspension angle may lead to a reduced height but this is not a good solution, A large angle at the trunk may cause whiplash or “bell ringing’ Pre-sternuim (frontal) attachment is not addressed in this study because such attachment is normally used with vertical arrester devices with rapid ‘snatch’ or arrest characteristics, hence relatively short falls. In these circumstances the frontal attachment may be the most suitable solution (e.g. for ladder climbing) but a fall may expose the user to facial injury from connectors, etc. f b) c) x NYA @ Vy )° LAPS Cer Figure 7. Illustration of harness position a) before fall, b) after fall, c) badly fitted 165.4 COMBINATION OF FACTORS AFFECTING FALL-ARREST HEIGHT The primary objective of this study is to seek to improve safety for workers who, on “low roofs", may have no other anchorage than the beam or structure at, or below, their feet. ‘The height from anchorage to floor or ground level may be less than the necessary deployment length of the fall safety system. For this reason the following comments relate to the usual “simple” safety system of connectors, lanyard, energy absorher and harness The factors affecting fall-arrest height are: * Position and type of anchorage. Selection of the anchorage is determined by conditions at the work site, ‘The anchorage may be an eyebolt or similar. It often is a wire-rope sling secured around a steelwork beam. ‘The length of the sling is important. ‘The author has seen slings in use that are too long for the girth of the beam, This adds to the drop height. * Choice of material and construction for the lanyard (webbing oF rope) affeets the extension fof the lanyard at peak arrest force. The ideal lanyard would be one that behaves ‘plastically’, ic. stretches to point of peak arrest foree and does not “rebound” when the peak force is passed. It should be noted that a fest house measures the extension length “post- test’, not the length at instant of peak force on a dynamic test ‘+ _Decelerative force of the energy absorber under dynamic conditions. © Choice of material and design of the energy absorber affects the extension of the energy absorber at peak arrest force. Energy absorbers by their very design are intended to behave as if “pl but usually there is some inherent elasticity due to the material of manufacture. Again, it should be noted that a test house measures the extension length “post-test”, not the length at instant of peak force on a dynamic test. * Selection of the hamess is a company oF individual choice. Proper adjustment of the harness is a very important individual skill © Extension of the harness at peak arrest. ‘This is acknowledged by some in the industry, but often the amount of ‘stretch’ is not known. The geometry ofa harmess alters considerably between “comfortable to work in’ and “suspended in’ (e.g. thigh straps which are near “horizontal” for working become ‘near vertical’ when the wearer is suspended in the harness {see Figure 7b). Measured from floor to harness-lanyard attachment point, many harnesses at suspension of the wearer increase that height by 200-300mm. If the attachment “Dring js hanging downward prior to suspension a further 100mm may be added to the height (most “D" rings are approximately SOmm across the half-circle"), ie. the apparent “stretch? becomes 300-00mm, * Angle of the spine at peak arrest force (not measured on test: it should be noted that the angle is measured ‘post test). The position of the lanyard attachment point on the harness is important. Figure 7c shows the effect a badly fitted hamess has on the angle of suspension. Such an angle at peak force would expose the wearer to possible whiplash or *bell-ringin, injury in a fall event.5.5 OBSERVATIONS ON FALL-ARREST CALCULATIONS In the era of rope-only lanyards for 2m, FF 1.0 falls (e.g. BS1397:1979), it was usual to consider factors for fall-arrest forces comparing performance with torso or anthropometric dummies. This, was advisable because the torso dummy behaved as a solid mass whilst the anthropometric dummy. behaved as a series of hinged masses (as does the human body), the latter resulting in lower peak arrest forces. With the advent of tear-ply and similar energy absorbers with a ‘flat peak” arrest forve, such factors are of litle consequence because the energy absorber behaves as a ‘Force limiter The combined performance of the lanyard type, energy absorber design and hamess ‘stretch’ is, usually the major influence in the performance of the fall-arrest system, These are considered at length in Annex B and Tables 2-10 therein, but the trends can be summarised. These are: © Deceleration distances tend to obey mathematical laws, As the deceleration force is, increased, the distance travelled reduces as a geometric series, For example, in the ematies expression vi=2x fxs, where v= velocity, f= acceleration, s = distance as °f increases, °s’ decreases. geometrically. Thus, if we consider s = 2 when f= 2, then s = [when f= 4, s = 0.667 when f = 6, s = 0.5 10 when f= 8. The reduetion in distance travelled poe isa “diminishing return’. (Some deviation from 6 this mathematical rule takes place in individual fo4 eases due to the varying performance of eo webbing, rope, tear-ply and hamess ‘stretch’, hence it was deemed advisable by the author to t 2 draw up the examples shown in Annex.‘ stance S Tables 2-10). ion of a nylon rope type lanyard is a very suitable “low jolt’ solution where there is no Fisk of the worker hitting the structure in fall. ts use ina 2m lanyardienengy absorber assembly may add 300-400mm to the overall fall height * Existing standards consider only one body wei KN results in a 6G arrest at this body weight ght, 100ky. An energy absorber operating at © Where a worker of body weight greater than 100kg falls on a 6kN nergy absorber there is a risk that the energy absorption will be inadequate for the kinetic energy of the fall, and the worker would be exposed to arrest at much greater than 6.1G (see Annex 'B', Figure 8, page 29), + Ifthe body weight of the worker is substantially less than 100kg the worker will be exposed to arrest deceleration greater than 6G when using a 6kN energy absorber, ght (rather than a 6kN 1m to the overall fall- + The effect of using a 4KN energy absorber with a 100kg body w device) will result in a 4G peak arrest, but this can add as much as height ina PF 2.0 fall (compare Table 4 with Table 7) © Any case for considering an SkN energy absorber for a 4m Fall (FF 2.0) of a L00kg person is weakened on learning that this would reduce the overall fall height by only approximately 0.175 m (when compared with a 6kN device — Annex 'B', compare Table 7 with Table 10). 186 CONCLUSIONS AND RECOMMENDATIONS Seated, and suitably constrained, the human torso ean withstand positive G 'forces* up to 40G for very short duration, ‘The literature indicates 40G up to S0ms (0.055). At such short duration the amplitude, or distance of arrest, is extremely small, Beyond these levels of force and duration there is considerable risk of severe injury. The literature also describes survived falls “from great heights’ into snow [Lancaster rear gunner] and paddy fields, even deliberately ‘wrapped in straw bundles’ and dropped into snow (Delahaye [13], Smuder 40]). but these are outwith the purpose of this study. Ejection seats have been in use since the 1940s with ejection forces, measured at seat level, of 18G to 25G and duration up to $00ms (0.53). Constraint during ejection is importa including head and neck. Recent designs, Martin Baker ME16 and US models, aim at to lessen risk of injury. (Note: television Channel 5 “When Pilots Ejeet”. 27 January 200: documented the survival story of a mic-air crash of two Russian MIG 29 fighters. with ‘model K36D ejection seats with a claimed 12G acceleration), In a parachute harness, the literature supports the argument that the human torso ean safely withstand fall-arrest deceleration of 12G (US military 5% risk of injury level is 12.1G). Straps must be at the shoulders. This configuration provides optimum alignment of spine. The “arbitrary” selection of OKN limit for wearers of industrial harness appears to have been a justifiable “interim” solution in view of the preferred dorsal attachment position and its necessary deviation from “optimum spinal alignment’. The 6kN limit applies to a body ‘weight of 100kg, ie. 6G maximum deceleration The medica/biomechanical literature reviewed in this study does not support an increase in arrest deceleration beyond 6G for wearers of industrial hamess (though it is known that users of low body weight may be exposed to +G, arrest deceleration ureater than 6G = see 7, below). In addition to the increased risk of injury. the study indicates that possible reduction in fall-arrest distance derived from increased arrest force is a “diminishing return’ It is the opinion of the author that the 6kN maximum ‘one-size” energy absorber is inadequate for the wide range of body weight in industry To date, the limit of GKN has been related to test dummies (and assumed human body weight) of 100kg. It would appear more practical to apply the OkN limit to a range of body weight. It is suggested in this study that the range should be 80kg to 100kg. With an energy absorber operating up to OkN, an SOky person would experience a maximum of 7.4G whilst a 100kg person would experience a maximum of 6G (but note that such absorbers ‘in fact’ usually are designed to operate at ARN to 4.5kN) An energy absorber of kN limit would be more suitable for body weights in the range 50~ 80kg. At this force a SOkg person would experience 8G maximum whilst the 80kg person would experience 5G maximum. The present 6KN limit for energy absorbers eould give rise to accelerations that may be injurious to persons of low body weight (a SOkg person could experience 12G), For body weights in the range 100-140kg itis recommended that an energy absorber limit of | SKN would be more suitable. At this force a 100kg person would experience 7.9G maximum whilst the 140kz person would experience 5.7G maximum. ‘The present 6kN limit for energy absorbers exposes persons of high body weight to ‘run-out’ of the device and consequent, possibly injurious, arrest forees when the absorber runs out of capacity (see Figure 8, Annes 'B' page 29), pull-out force of 4KN to 4.SKN increases this risk. 19