辽宁太子河流域非点源污染模拟研究--《中国农业大学》2013年博士论文
辽宁太子河流域非点源污染模拟研究
【摘要】：随着社会经济的快速发展,辽宁太子河流域社会环境、流域河流污染问题日益突出,各类工农业污染源带来的有机污染、重金属、农药等污染物给流域水环境带来重大压力。随着太子河流域对点源污染控制能力的提高和技术的相对成熟,非点源污染已成为影响水环境恶化的主要来源。因此,针对非点源污染特征,开展污染负荷定量化研究,识别污染物来源和关键源区,研究不同水文条件、不同土地利用情况、水土保持措施和农业管理措施等对流域非点源污染负荷的影响,是进行水环境质量评估、风险评价及污染物削减的重要依据,也是有效控制水环境污染、保障用水安全的重要手段。本论文的主要研究内容及结论如下：
(1)在详细的实地调查、资料收集和模型比选的基础上,基于SWAT模型(Soil and Water Assessment Tool)建立了太子河流域的分布式非点源污染模型。应用LH-OAT (Latin-Hypercube One-factor-At-a-Time)方法进行了参数的灵敏性分析,筛选并确定了模型中需要率定的相关参数,采用SCE-UA方法分别以河道径流量、输沙量和水质指标为目标变量率定相关参数,确定了模型水文、泥沙和水质参数值。率定期R2均大于0.58,NS大于0.60,验证期R2大于0.45,除南甸站以外NS均大于0.50,河道径流模拟效果较好,泥沙和水质模拟结果符合要求,表明模型在太子河流域适用性较好。
(2)开展了太子河流域降水量、地表产水量、泥沙流失量和非点源污染物(总氮TN和总磷TP)入河量的时空动态分析,讨论不同水文年型不同降雨情况下,现状年(2008年)非点源污染物的时空变化规律,及土壤侵蚀的空间分布情况,识别污染关键源区。结果表明：年流域年均降水量为92.17亿m3,地表产水量为31.35亿m3,泥沙流失量为229.50万t,TN负荷为17357.43t,TP负荷为7110.91t,流域降水量呈东南向西北减弱的趋势,最大值发生在降水量较大的2005年,TN和TP负荷分别为23030.00t、9346.35t。通过多年模型模拟结果,基本上可以确定太子河流域内非点源污染负荷关键源区是下游地区的灯塔市和辽阳县,氮磷负荷强度空间差异较大,全流域多年平均为13.20kg/hm2和5.41kg/hm2。氮磷负荷的时空分布与降水量关系较大,汛期(6-9月)负荷总量占全年的77.76%和80.00%。
(3)分析了现状(2008年)经济、人口、气象等水平下,改进施肥方式(情景1)、减少50%施肥量(情景2)、采用水土保持措施(情景3)、增加城镇地区透水面积(情景4)和设置河岸缓冲带(情景5)不同措施对非点源污染负荷的控制效果。5种情景下泥沙流失总量分别减少1.69%、1.54%、33.84%、4.28%、29.62%,TN负荷总量减少7.41%、28.1%、24.38%、2.79%、28.0%,TP负荷总量减少11.39%、41.77%、37.93%、3.36%、36.49%,现状年的TN平均负荷强度为13.72kg/hm2,5种情景中平均负荷强度分别减少1.02kg/hm2、3.86kg/hm2、3.35kg/hm2、0.29kg/hm2、3.84kg/hm2;现状年的TP平均负荷强度为5.52kg/hm2,5种情景中分别平均负荷强度分别减少0.63kg/hm2.2.30kg/hm2、2.09kg/hm2、0.18kg/hm2.2.01kg/hm2,不同土地利用类型的单位面积负荷均不同程度的减少,情景2、3、5有较好的污染控制效果。
(4)采用全国1km网格土地利用数据库(,2000)数据,分析了太子河流域在80年代、90年代土地利用情况的动态变化过程,并进一步研究了土地利用／覆盖变化(land useand land cover change, LUCC)对流域水文水资源和非点源污染的影响。年太子河流域内地类转化主要是在水田、旱地、疏林地、城镇用地和农村居民点用地之间,80年代至90年代中期土地利用变化比90年代后期变化要剧烈。1980s到2000年水田和疏林地减少40.81km2、112.34km2,旱地、农村居民点用地、城镇用地、工交建设用地分别增加113.30km2、16.73km2、19.50km2、10.04km2,水田转化为农村居民点用地、城镇用地和工交建设用地,而疏林地转变为早地。1980s土地利用情况下,年多年平均地表产水量、泥沙流失量、TN和TP入河量分别为33.30亿m3、207.80万t、16756.68t和95年土地利用情况下平均值分别为32.83亿m3、210.74万t、16917.52t和6643.84t,空间差异较大。与2000年土地利用情况下相比,1980s和1995年土地利用情况下地表产水量变化较大区域为观音阁水库上游、观音阁水库库区和下游海城地区,不同子流域的TN、TP污染负荷强度有增有减,总体上是负荷强度高的区域变化率较小。
【关键词】：
【学位授予单位】：中国农业大学【学位级别】：博士【学位授予年份】：2013【分类号】：X522【目录】：
摘要5-7Abstract7-11插图目录11-13附表目录13-15第一章 绪论15-28 1.1 研究背景及选题意义15-17 1.2 国内外研究进展17-25 1.3 研究内容与方法25-28第二章 材料与方法28-55 2.1 研究区概况28-34
2.1.1 地质地貌、土壤植被30-31
2.1.2 气象水文31-32
2.1.3 水利工程32-33
2.1.4 社会经济33-34 2.2 SWAT模型介绍34-40
2.2.1 模型概述34-36
2.2.2 计算原理36-40 2.3 模型输入及构建40-50 2.4 参数敏感性分析及模型率定与验证方法50-55第三章 参数敏感性分析及率定验证55-73 3.1 敏感性分析55-58
3.1.1 径流量55-56
3.1.2 泥沙量56-57
3.1.3 营养盐57-58 3.2 率定与验证结果分析58-72
3.2.1 径流量率定验证结果58-67
3.2.2 泥沙量率定验证结果67-69
3.2.3 水质率定验证结果69-72 3.3 小结72-73第四章 流域非点源污染负荷特征分析73-86 4.1 非点源污染负荷时空分布规律分析73-78 4.2 不同水文年污染负荷特征分析78-83
4.2.1 流域水文条件78-79
4.2.2 负荷时空分布分析79-83 4.3 土壤侵蚀关键区识别83-84 4.4 小结84-86第五章 污染控制情景下非点源污染负荷特征分析86-95 5.1 情景设置86-87 5.2 不同污控情景下污染负荷特征分析87-90 5.3 不同情景下单位负荷变化分析90-94 5.4 小结94-95第六章 土地利用变化对非点源污染负荷的影响分析95-107 6.1 土地利用变化分析95-101 6.2 土地利用变化对污染负荷的影响101-106 6.3 小结106-107第七章 结论与建议107-110 7.1 主要结论107-108 7.2 创新点108-109 7.3 不足及建议109-110参考文献110-121致谢121-123作者简介123
欢迎：、、)
支持CAJ、PDF文件格式
【参考文献】
中国期刊全文数据库
陈军锋,陈秀万;[J];北京大学学报(自然科学版);2004年02期
夏智宏;周月华;许红梅;;[J];长江流域资源与环境;2010年02期
胡连伍;王学军;罗定贵;蒋颖;;[J];地理与地理信息科学;2006年02期
杨桂莲,郝芳华,刘昌明,张雪松;[J];地理科学进展;2003年05期
王秀兰,包玉海;[J];地理科学进展;1999年01期
王思远,刘纪远,张增祥,周全斌,赵晓丽;[J];地理学报;2001年06期
黄满湘,章申,张国梁,张秀梅;[J];地理学报;2003年01期
李跃勋;徐晓梅;何佳;郑一新;张琨玲;陈云波;李中杰;;[J];湖泊科学;2010年05期
陈友媛,惠二青,金春姬,邱汉学,吴德星;[J];环境科学研究;2003年01期
胡远安,程声通,贾海峰;[J];环境科学研究;2003年05期
中国博士学位论文全文数据库
沈晔娜;[D];浙江大学;2010年
【共引文献】
中国期刊全文数据库
唐丽娟;雷磊;曹玲;;[J];Asian Agricultural R2010年01期
;[J];Asian Agricultural R2011年10期
;[J];Asian Agricultural R2012年01期
张秀玲;张保华;蒋玉华;薛杰文;;[J];山东农业科学;2011年02期
窦培谦,王晓燕,秦福来,王丽华;[J];安徽农学通报;2005年04期
董林林;韦良焕;;[J];安徽农学通报(上半月刊);2009年05期
崔颖;高敏华;;[J];安徽农学通报(上半月刊);2009年05期
王桂苓;马友华;石润圭;黄文星;孙兴旺;吴春蕾;;[J];安徽农学通报(上半月刊);2009年07期
张竹华;陈龙乾;王秉义;张奎;;[J];安徽农学通报(上半月刊);2010年07期
李扬;;[J];安徽农学通报(上半月刊);2010年09期
中国重要会议论文全文数据库
肖梦华;俞双恩;丁继辉;;[A];土肥水资源高效利用与农业面源污染防控技术研讨会论文集[C];2011年
任秀文;姜国强;钟满妮;李开明;;[A];2011中国环境科学学会学术年会论文集（第一卷）[C];2011年
陈泓熹;;[A];2011中国环境科学学会学术年会论文集（第三卷）[C];2011年
;[A];Proceedings of Symposium from Cross-strait Environment & Resources and 2nd Representative Conference of Chinese Environmental Resources & Ecological Conservation Society[C];2010年
周堂;石培基;;[A];黄河流域资源环境与生态文明建设学术研讨会交流材料[C];2011年
李雪莹;;[A];转型与重构——2011中国城市规划年会论文集[C];2011年
杜加强;王金生;舒俭民;滕彦国;;[A];第二届生态补偿机制建设与政策设计高级研讨会论文集[C];2008年
陈刚;邓文胜;王丽亚;;[A];全国国土资源与环境遥感应用技术研讨会论文集[C];2009年
沈涛;党安荣;;[A];中国测绘学会第九次全国会员代表大会暨学会成立50周年纪念大会论文集[C];2009年
许继军;杨大文;雷志栋;;[A];节能环保 和谐发展——2007中国科协年会论文集（二）[C];2007年
中国博士学位论文全文数据库
李杰;[D];中国海洋大学;2010年
武桂芝;[D];中国海洋大学;2010年
余云军;[D];中国海洋大学;2010年
王雄师;[D];兰州大学;2010年
穆振侠;[D];新疆农业大学;2010年
沈晔娜;[D];浙江大学;2010年
胡军强;[D];浙江大学;2009年
曾和平;[D];昆明理工大学;2009年
彭保发;[D];中南林业科技大学;2010年
欧洋;[D];首都师范大学;2011年
中国硕士学位论文全文数据库
毕磊;[D];华中农业大学;2010年
于波;[D];华中农业大学;2010年
郝慧君;[D];华中农业大学;2010年
艾蕾;[D];华中农业大学;2010年
杜金龙;[D];华中农业大学;2010年
李承力;[D];华中农业大学;2010年
雷征;[D];华中农业大学;2010年
袁昭;[D];华中农业大学;2010年
曾繁富;[D];河南理工大学;2010年
刘丽娜;[D];郑州大学;2010年
【二级参考文献】
中国期刊全文数据库
刘萍萍,闫艳春;[J];山东农业科学;2005年02期
汤进华;;[J];安徽农业科学;2006年12期
冯庆;王晓燕;王连荣;;[J];安徽农业科学;2009年24期
陈文英,毛致伟,沈万斌,孙述海;[J];北方环境;2005年02期
李海雯;陈振楼;王军;史贵涛;;[J];环境科学与管理;2007年03期
陈军锋,陈秀万;[J];北京大学学报(自然科学版);2004年02期
王海龙;余新晓;武思宏;张晓明;谢媛媛;;[J];北京林业大学学报;2007年S2期
郑凡东;孟庆义;王培京;金桂琴;;[J];北京水务;2007年05期
丁晓雯;刘瑞民;沈珍瑶;;[J];北京师范大学学报(自然科学版);2006年05期
徐清,杨天行,刘晓端,葛晓立;[J];吉林大学学报(地球科学版);2003年03期
中国重要会议论文全文数据库
朱建国;郭红岩;王晓蓉;;[A];中国土壤学会第十次全国会员代表大会暨第五届海峡两岸土壤肥料学术交流研讨会论文集（面向农业与环境的土壤科学综述篇）[C];2004年
中国博士学位论文全文数据库
李硕;[D];南京师范大学;2002年
吕唤春;[D];浙江大学;2002年
庞靖鹏;[D];北京师范大学;2007年
张超;[D];清华大学;2008年
中国硕士学位论文全文数据库
王晓峰;[D];首都师范大学;2001年
庄咏涛;[D];西安理工大学;2002年
代俊峰;[D];华中农业大学;2004年
秦福来;[D];首都师范大学;2006年
曹彦龙;[D];重庆大学;2006年
王晓燕;[D];河海大学;2007年
刘明宇;[D];首都师范大学;2008年
孟祥明;[D];天津大学;2007年
【相似文献】
中国期刊全文数据库
杨文龙，杨树华;[J];云南环境科学;1996年03期
汪雅谷;[J];上海农业科技;1996年06期
洪小康,李怀恩;[J];西安理工大学学报;2000年04期
李贵宝,李建国,毛战坡,尹澄清,王为东;[J];南水北调与水利科技;2005年01期
William·F·R蒋林燕;王秉忠;;[J];农业环境与发展;1990年03期
辜来章;吴德玲;郝淑英;;[J];云南环境科学;1991年01期
李怀恩,沈晋,刘玉生;[J];环境科学学报;1997年02期
路月仙,陈振楼,王军,许世远;[J];环境保护;2003年11期
许书军,魏世强,谢德体;[J];长江流域资源与环境;2004年04期
杨建云;[J];云南环境科学;2004年S1期
中国重要会议论文全文数据库
谈俊益;邵孝候;吴俊峰;陈丽娜;李圆圆;文涛;;[A];2010中国环境科学学会学术年会论文集(第二卷)[C];2010年
陈泓熹;;[A];2011中国环境科学学会学术年会论文集（第三卷）[C];2011年
赵同科;张强;;[A];全国农业面源污染与综合防治学术研讨会论文集[C];2004年
朱建国;郭红岩;王晓蓉;;[A];中国土壤学会第十次全国会员代表大会暨第五届海峡两岸土壤肥料学术交流研讨会论文集（面向农业与环境的土壤科学综述篇）[C];2004年
张春玲;;[A];中国水利学会2002学术年会论文集[C];2002年
安沫平;刘峰;;[A];全国农业面源污染与综合防治学术研讨会论文集[C];2004年
王晓燕;;[A];全国农业面源污染与综合防治学术研讨会论文集[C];2004年
杨柳;马克明;陈利顶;;[A];“土地变化科学与生态建设”学术研讨会论文集[C];2004年
林昭远;陈键鑫;颜正平;;[A];海峡两岸减轻灾害与可持续发展论文专辑[C];2000年
冷罗生;;[A];环境法治与建设和谐社会——2007年全国环境资源法学研讨会（年会）论文集（第三册）[C];2007年
中国重要报纸全文数据库
李红梅;[N];中国水利报;2009年
记者 张岩;[N];黄河报;2011年
李洁尉;[N];广东科技报;2002年
陈云芬;[N];云南日报;2009年
操秀英;[N];科技日报;2007年
王耀兵;[N];中国渔业报;2005年
民革长春市委员会 民建长春市委员会 长春市工商业联合会 民盟长春市委员会 民进长春市委员会 九三学社长春市委员会 农工民主党长春市委员会;[N];长春日报;2008年
蔡明（西安理工大学环境科学研究所）;[N];中国水利报;2006年
记者 杨亚非 特约通讯员 李德旺 王孟;[N];人民长江报;2008年
本报记者 黄里;[N];四川日报;2005年
中国博士学位论文全文数据库
罗倩;[D];中国农业大学;2013年
吴磊;[D];重庆大学;2012年
孟凡德;[D];首都师范大学;2013年
韩成伟;[D];大连理工大学;2012年
李伟;[D];浙江大学;2013年
沈晔娜;[D];浙江大学;2010年
史淑娟;[D];西安理工大学;2010年
李爽;[D];山东师范大学;2012年
吕唤春;[D];浙江大学;2002年
欧洋;[D];首都师范大学;2011年
中国硕士学位论文全文数据库
王照蒸;[D];首都师范大学;2004年
孙莉宁;[D];合肥工业大学;2005年
谷丰;[D];浙江大学;2005年
毛致伟;[D];吉林大学;2005年
张军;[D];东北师范大学;2006年
邵兰霞;[D];东北师范大学;2006年
梁博;[D];首都师范大学;2005年
黄宗楚;[D];华东师范大学;2005年
钱德安;[D];吉林大学;2006年
王晓辉;[D];合肥工业大学;2006年
&快捷付款方式
&订购知网充值卡
400-819-9993
《中国学术期刊（光盘版）》电子杂志社有限公司
同方知网数字出版技术股份有限公司
地址：北京清华大学 84-48信箱 大众知识服务
出版物经营许可证 新出发京批字第直0595号
订购热线：400-819-82499
服务热线：010--
在线咨询：
传真：010-
京公网安备75号PMCID: PMC3400220Seated Weight Distribution of Adults and Children in Normal and Non-Normal Positions, , , and
This study investigates occupant’s body weight distribution on various seat components at different seated positions. A test seat with instrument panel and door prototypes was constructed, and load cells used to quantify load distribution on seat and other interior components. Three experiments were conducted, the first used 36 adult subjects at 13 selected normal positions, the second with 11 adults at 17 non-normal positions, and the third with 19 children at seven positions. For adults, the lowest load on seat cushion is 51.1% of body weight for normal seated positions, and 55.1% for non-normal positions. The lowest load on seat cushion and back is 76% of body weight for normal positions, and 55% for non-normal positions. For children at various positions, the highest load to seat cushion is 93.4%, and to the cushion and back 94%. With 2% of all tests, the cushion load is less than 30kg, and cushion and back summed load less than 40kg. Using cushion based occupant detection technology, there is an overlapping zone between small female adults and 30kg children. The frame-based technology can increase discrimination power. Furthermore, reducing critical weight for airbag suppression from 30kg to 24kg can significantly minimize, even eliminate, the overlapping zone.The objective of the first experiment was to investigate adult’s body weight distribution on seat cushion, back, footrest and armrest with various cushion angle, back angle, and seat height at normal sitting position.METHODSTest Setup The test seat consists of four components: back, cushion, footrest and armrest (), each component independent of others for load bearing. Cushion angle is adjustable from 0° to 16°. Four load cells, each with a capacity of 2000 lbs and an accuracy of ±2%, are mounted under the corners of cushion frame at 480 mm apart laterally and 400 mm anterior-posterior. Back angle is adjustable from 0°, 12°, 24°, 36°, to 45°. Footrest is adjustable to a footrest-cushion height of 216, 322, and 428 mm, and the two load cells, each with a 50 lbs capacity, are 304mm apart under footrest board. Two armrest load cells of 50 lbs capacity are 708 mm apart and 250 mm above the top of the cushion at 0°.Test setup for experiment 1.SubjectsThirty-six adult human subjects, 18 females and 18 males, participated in this experiment. Subjects, as summarized in , are categorized by body weight and gender into small, medium, large females (SF, MF, and LF), and small, medium, and large males (SM, MM, and LM).Subjects’ anthropometric data for Experiment 1.Experimental Design Three design variables are cushion angle (8°, 12°, 16°), back angle (12°, 24°, 36°), and seat height (L=216, M=322, H=428 mm). Control variables are armrest usage, and foot extension. When feet are not extended, subject’s heel points are 150 mm forward from cushion front edge.The experiment is arranged according to Box-Behnken design in order to reduce the number of tests and to simulate a response surface design for identifying factor level combinations for minimum cushion load. Conditions 13 to 15 are repeated tests ().Test arrangement for Experiment 1.Data Collection Procedure Each subject filled out an informed consent form before test. His/her body weight was recorded. The measurement system was calibrated before tests and after each test. Each condition was tested twice while the subject was seated: first with feet close to cushion edge and second with feet being extended as much as the subject felt comfortable.RESULTSUnder each test condition, measurements include load on cushion and the calculated load sum on seat cushion and back. Each data set is presented as load magnitude in kg and proportion of body weight in percentage.Results have shown that feet extension actually causes more cushion load, which is favorable for occupant discrimination, analysis will focus on tests with feet close to seat cushion.
depict confidence intervals of the estimated means.Confidence intervals of mean cushion load.Confidence intervals of mean cushion and back load.Cushion Load Among 15 test conditions, cushion load for each subject group is the lowest under Test No.6 with back angle 36°, cushion angle 12°, and seat height low (36°/12°/216mm). Mean cushion load for SF group is 30.3±1.7kg (M±SD), or 56.1±4.2% of body weight. Between-group difference in cushion load is obvious. Percentage-wise however, between-group difference is very small and non-systematic.Cushion and Back Load The sum of cushion and back load is apparently larger than cushion load only. The lowest load on cushion and back is under Test No.9 (24°/8°/216mm). For small female group, the lowest load is 42.8±2.0kg, or 79.2±2.5% of body weight. For all groups, load sums are larger than 76.6±1.9% of body weight.Analysis of variance (ANOVA) reveals that cushion load is significantly affected by all three factors, F(1, 528)=4.12, p<.05 for back angle, F(1,528)=6.72, p&#x for seat height, and F(1, 528)=3.97, p&#x for cushion angle squared. Similar significance levels are found for their effects on cushion load percentage. None of the three factors has significant effect on the sum of cushion and back load. Gender has significant effects on both cushion load and the sum of cushion and back load. Weight group has significant effect on cushion load but not on the sum of cushion and back load. The degree of effect on percentage value of either measure is much smaller than on its load magnitude.Regression analysis was conducted for cushion load only. To consider between-subjects differences, subject group is included as a dummy variable in the model. The regression model of cushion load is (R2 = 0.842):Cushion_Load (kg)=         Cst_Sub_Grp-0.272*Bk_Angle+0.0738*St_Hgt+0.0247*Sq(Ch_Angle)Where Cst_Sub_Grp is 18.7kg for SF, 28.6kg for MF, 35.7kg for LF, 23.5 kg for SM, 31.9kg for MM, and 43.4kg for LM.Since cushion loading in percentage is more consistent among subject groups, a regression model of load percentage can be established without introducing the dummy variable for subject groups (R2 = 0.772):Cushion_Load (%)=41.7         -0.378*Bk_Angle+0.105*St_Hgt+0.036*Sq(Ch_Angle)DISCUSSIONThe objective of this study was to investigate adult’s body weight distribution on seat components at various cushion angles, back angles, and seat height. The cushion load cells showed a −4% systematic error because of their large load bearing capacity of 2000 lbs each.Test condition No.6 with the lowest cushion load of 56% was identified. For a 5th percentile female adult weighing 46kg (101 lbs), the lowest cushion load would be 25.7kg. Using the regression model for cushion load percentages, the worse case within the design space is 36°/8°/216mm. The predicted cushion load would be 51.1% or 23.5kg for a 5th percentile female.The fact that actual cushion load from an adult can be below 30kg adds uncertainty to occupant discrimination based on cushion load. However, the sum of cushion and back load is always larger than cushion load and it is less sensitive to seat adjustment. Under no test conditions was the sum of cushion and back load less than 42kg. ANOVA showed that the summed load is not significantly affected by cushion and back angles, because change in angles only redistributes load between seat cushion and back. The only loss of measured body weight is through armrest and footrest. Assuming footrest load is less than 26% and armrest load less than 8%, the sum of cushion and back load should be no less than 66% of the body weight. With a small subject of 46kg, the load sum is no less than 30.4kg.This experiment was carried out under controlled conditions and it does not address seated load for non-normal positions. Investigation of load distribution with those test conditions is the objective of the other two experiments.The second experiment was aimed to determine adults’ seated body weight distribution in non-normal positions. By non-normal we mean that an occupant does not assume largely symmetrical postures and does not make effective use of seat cushion and back. The weight distribution of the subject is assessed through the seat cushion, back, footrest, armrest, instrument panel (IP) and door.METHODSTest Setup Wooden prototypes of a height adjustable IP and a side door were mounted to meet the packaging relationships for a GM Pontiac mid-size passenger car. Test setup is similar to that in Experiment 1 and load cells were used to measure the load applied to the seat-cushion, footrest, armrest and IP. The seat back, door and the IP were not instrumented, with the exception of Position 17 at which the footrest load cells were used to measure load taken by the IP.Subjects Eleven adult human subjects, who had attended the first part of the study, participated in this experiment. The subjects were divided into female and male groups ().Subject data for Experiment 2.Test Arrangement The principal factor of this part of the study is the various seated positions. The positions shown in
were chosen based on literature, various testing protocols and on an in-house survey. Those positions represent typical non-normal sitting positions that may be assumed by vehicle occupants. Under all positions but Position 4, the cushion was set to 0° and the back reclined to 24°. In Position 4, seat back was reclined to 45°.Description of non-normal positions in Experiment 2.Measurements and Analysis The test procedure is similar to that of Experiment 1. Load distributed through the seatback, door or IP was determined by subtracting the weight of the subject and the load applied to the seat-cushion, armrest and footrest. There were no positions where the subject was loading both the IP and footrest.RESULTSExperiment 2 leads to the following findings:Absolute load levels do not represent applicable data because the mixed subject group and the relatively small sample size. Percentage load data allows a more consistent comparison across different subject groups.With all subjects under all 17 positions, the measured load taken by each seat and interior components varies as following:-
•Cushion only:55.1±4.5% to 95.4±1.8%•Cushion & back summed:55.1±4.5% to 99.8±0.2%•Footrest or floor:13.1±6.1% to 32.0±4.0%•Instrument panel from arms:14.7±2.8% to 16.5±3.5%•Door from arm:14.5 ± 2.6%•Armrest:5.1±2.2% to 8.8±3.0%The lowest cushion load, 55.1±4.5%, occurred when subjects were leaning forward to the IP at Positions 10 and 11 when arms and head were on the IP or the seat back was reclined to 45°. Other positions with low cushion load were those when subjects leaned forward reaching HVAC/audio controls and glove compartment at Positions 7 and 9.Generally the sum of cushion and back load is larger than cushion load only. The exception is Position 10 at which the seat back was not loaded and the cushion and back summed load was 55.1±4.5%, the same as cushion load only. The next lowest cushion and back summed load is 67.1 ± 5.2% at Position 5.This part of the study was carried out to determine the load applied to seat components when subjects sit in 17 non-normal positions. Load applied to the seat cushion varied with occupant positions, with the average load of 55.1±4.5% to 95.4±1.8% of body weight. The minimum cushion load occurred in Positions 10 and 11.While all the 17 positions are somehow non-normal, they may not be out of position. The positions are non-normal in that with those positions, subjects do not effectively use the seat cushion and back support. At positions under which the seat cushion and back are effectively used, cushion load will be similar to that under normal positions as shown in Experiment 1, although the arms may rest on the door or sitting postures asymmetrical. Usually asymmetrical postures affect seated pressure distribution on the occupant-seat interface, but they do not necessarily influence load distribution as measured on the entire contact area.The difference in cushion load and the summed load from cushion and back confirms the findings of Experiment 1. The summed load provides a better safe margin for detection of small adults.The third experiment was to determine the weight distribution when children sit in various positions. Weight distribution was assessed through the seat cushion, back, footrest and IP. The major interest is to identify the maximum load on cushion and the maximum summed load on cushion and back with children subjects.METHODTest Set-Up The set-up is the same as that in Experiment 2.Subjects Nineteen child subjects were tested, 11 males and 8 females, ages between 2.5 and 12 years old. Subjects are divided into two groups, that less than 30kg and that more than 30kg of their body weight. Average weight is 19.6±4.6kg for the light group, and 39.2±8.4 kg for the heavy group.Experimental Arrangement Seat cushion was set to 0° and the back was reclined to 24° except for Position 4 when seatback was reclined to 45°. The sitting positions used in this study were:Position nDescription1Leaning forward, feet touching footrest2Leaning forward, feet hanging3Sitting on the knees facing backward, leaning on the back4Sitting with two legs crossed under5Facing backward, placing stomach on seat cushion and knees on footrest6Sitting forward, leaning on IP7Kneeling on the seat and facing forwardMeasurement and Analysis The load distributed on the seat back or IP was determined by subtracting the weight of the subject from the load on seat cushion and footrest. There were no positions where the subject was loading both the IP and footrest.RESULTS shows the load distributed on the seat cushion, and on cushion and back. For both measurements at a number of positions, maximum mean load is from 91.6±3.0% to 94±0.1% of the body weight of the subjects. The small difference between cushion load and the summed load on cushion and back is due to the little loading on seat back by children.Seated load for children at seven positions.This study investigates the weight distribution on the seat cushion and on cushion and back for adults and children sitting in normal and non-normal positions. The study consists of three experiments. In the first, weight distribution of 36 adults sitting in normal positions was assessed at various seat cushion and back angles and seat height. In the second, weight distribution of 11 adults sitting in 17 non-normal positions was determined. In the third, 19 children were tested in seven positions.Current FMVSS 208 () requires that airbags be switched off if the occupant weighs less than 30kg although in the future the critical weight of 30kg may be reduced to 24kg (). A major concern is how well an occupant detection system can discriminate between adults and children. The first two experiments with adults show that whether it is normal position or non-normal position, the lowest load to cushion due to body weight is 51.1% of body weight. For a 5th percentile female of 46kg in body weight, the induced cushion load will be 23.5kg on average. In comparison, a child with shorter stature will load the seat cushion as much as 93% of his/her body weight. That is 27.9kg for a child of 30kg. Clearly we see an overlapping zone in weight discrimination which adds much difficulty to the reliability of cushion based occupant detection system. There are however a few aspects we can consider for a better solution.One possibility is to use seat frame based detection technology instead of the cushion based. With frame based technology, the measured load will be due to the sum of cushion and back load, thus potentially reduce the overlapping zone in weight discrimination.
provides scatterplots of seat load against body weight for all subjects, adults and children, at all positions tested. For cushion load only, most of the load due to adults’ body weight is above 30kg, but the sum of cushion and back load is mostly above 40kg. From , we can further see that for adult subjects, 0.9% of all the cases have cushion load less than 30kg, while 1.5% of the cases have the summed load less than 40kg. There is no case where the summed load is less than 30kg. Therefore, to use frame-based technology will significantly increase discrimination power.Scatterplots of load on seat vs. body weight.Histograms of load distribution on the seat.Another possibility is to reduce the critical weight level for airbag suppression. Since there is always possibilities of an overlapping zone in discriminating children from adults with either technology, a change in the critical weight from 30kg to 24kg can remarkably help minimize, or even eliminate, the overlapping. If using 24kg as the threshold, the maximum cushion load due to a child’s body weight will be about 22.3kg. This low level of cushion load creates a safe margin for occupant discrimination. On the other hand, the probability of less than 24kg cushion loading by a small adult is even smaller than that of 30kg.One thing that is not investigated in this study is the probability of assuming each of the tested positions. Analysis in previous text treats each seated position at an equal occurrence probability. If we could find, say, that the few critical non-normal positions at which cushion load is the lowest have a much smaller probability of occurrence in real situations, then cushion based technology becomes very feasible. Furthermore, one confounding factor may help increase decision power for airbag deployment or suppression although it does not increase occupant weight prediction power. When an adult of above 46kg in body weight is measured as being less than 30kg according to cushion load measurement, it may indicate a potential OOP situation, e.g. Position 10 in Experiment 2, and airbag should be deactivated.All the measurement and discussion in this work are for static situations. In a dynamic situation, total load and its distribution on the seat back and cushion can be very different. During slow speed rear impact for example (), seat back will be much more severely loaded due to occupant’s body weight and the kinetic energy. Accordingly, the amount of cushion load is reduced to a very low level at the peak of impact. Situations as such do not permit valid discrimination of occupants using weight based technology. In fact, occupant detection and discrimination should have to be made prior to any emergency situations such as pre-impact braking, bracing and actual impacts. This can be fulfilled by constant filtering of weight sensing signals during the entire process of vehicle movement, so that discrimination is not based just upon isolated peaks of signals.Measurements in this study were made using load cells on cushion and footrest. Load cells are accurate devices if appropriate capacity is chosen. Actual implementation of weight sensing technology may however be based on pneumatic or liquid bladder systems mounted on seat cushion. Calibration and validation of those systems are critical for those systems. Results of this study should be interpreted with regard to the specified load cell system.For adults, the lowest load on seat cushion is 51.1% of body weight for normal seated positions, and 55.1% for non-normal positions.For adults, the lowest load on seat cushion and back is 76% of body weight for normal positions, and 55.1% for non-normal positions.For adults, about 0.9% of all test cases induced a cushion load less than 30kg, and about 1.5% induced a cushion and back summed load less than 40kg.For children at various positions, the highest load to seat cushion is 93.4%, and 94% to the cushion and back.Using cushion based occupant detection technology, there is an overlapping zone between small female adults and 30kg children. The frame-based technology can increase discrimination power. Furthermore, reducing critical weight for airbag suppression from 30kg to 24kg can significantly minimize, even eliminate, the overlapping zone.De Leonardis DM, Ferguson
SA, Pantula JF. Survey of driver seating positions in relation to the steering wheel. SAE paper No. 98. National Highway Traffic Safety Administration. 49CFR571.208 Federal Motor Vehicle Safety Standards: Occupant Crash Protection. October 1, 1998.National Highway Traffic Safety Administration. 49CFR571, 587, and 595 Federal Motor Vehicle Safety Standards: Occupant Crash P Proposed Rule. Federal Register. 1998 September 18;63(181):4.Parkin S, Mackay GM, Cooper A. How drivers sit in cars. 37th Annual Proc. AAAM, Nov. 4–6; San Antonio, TX. 1993. Shen W, Wiklund K, Swamy B, Nilson G, Humer M. Pressure and load patterns on seat interface during low severity rear impact: A comparison between human subjects and crash dummies. Proc 1998 International IRCOBI Conf on the Biomechanics of I September 16–18; Chalmers, Sweden. 1998. pp. 453–464.Zacharkow D.
Posture: Sitting, Standing, Chair Design and Exercise. Charles C. Thomas P Springfield, IL: 1988. Articles from Annual Proceedings / Association for the Advancement of Automotive Medicine are provided here courtesy of Association for the Advancement of Automotive Medicine
)- subscription required
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
Show all items
CitePeer Related Articles}