基于复数傅里叶分析的岩土颗粒细观特征识别与随机重构方法

岩土细观特征的识别与分析是岩土力学研究的重要内容。基于傅里叶变换原理,将任意岩土颗粒的二维轮廓变换为傅里叶描述符与相位进行分析,建立了基于复数傅里叶分析的细观特征表征与重构方法。通过两组不同颗粒的傅里叶细观特征统计,研究了不同阶傅里叶描述符与细观特征参量的对应关系。结果表明：-12～0及16～20阶对应的傅里叶描述符决定颗粒的形状;-23～-11及1～15阶对应的傅里叶描述符与颗粒粗糙度密切联系;傅里叶描述符曲线峰值与颗粒的粒径显著线性相关;高阶傅里叶系数与阶数近似呈对数线性关系,低阶系数则由不同颗粒形状决定。采用建立的傅里叶描述符对应关系进行颗粒相似重构研究,发现重构颗粒的外轮廓及傅里叶统计参量与实际颗粒十分接近,且反映颗粒的粗糙度、纹理等细部特征,其成果可用于大量岩土颗粒与力学特性的相关性研究。     

动态图像法应用于海滩沉积物粒度粒形测试及其与筛析法的比较

筛析法是海滩沉积物粒度分析较经典和常用的方法。随着科学技术的发展,利用动态图像法分析沉积物粒度逐渐得到推广。本文利用动态图形法和筛析法对海南岛5个海滩剖面20组沉积物样品进行粒度粒形测试,并将两种方法所测得的粒度进行比较。测试分析结果显示,该方法的测试结果重复性好,精度高;通过与筛析法的对比显示,动态图像法与筛析法的测试结果非常接近,粒度级配曲线基本一致,各个粒度参数值很接近且相关性非常好（R2>0.94）;由动态图像法得出的粒形参数可以看出,粒径相当的不同海滩沉积物粒形参数有很大差别,同一海滩不同部位的球形度和宽长比变化很大,对称度和凹凸度变化稍小。研究表明,动态图像法与筛析法之间的粒度分析差异主要来自于两者测量原理的不同和天然海滩砂颗粒形状的不规则;动态图像法解决了不规则沉积物粒度的测量。因此,动态图像法可以替代筛析法来测量沉积物粒度,应用前景广阔。     

扬子陆块西缘寒武系砂岩的物源分析：对古地理位置重建和构造背景的指示

扬子陆块西缘寒武系主要为一套碎屑岩- 碳酸盐岩的岩石组合,前人研究多认为形成于相对稳定的克拉通盆地。但同时期出现的大陆岩浆作用显然与前期认定的克拉通盆地性质不符,需要借助扬子西缘的物质来源探讨构造背景。基于野外露头等资料,本文通过对扬子陆块西缘会泽和会东附近寒武系3件砂岩样品进行重矿物分析、电气石电子探针和碎屑锆石U- Pb测年分析,确定扬子西缘寒武纪沉积物的源区;并结合沉积序列等综合探讨扬子陆块西缘寒武纪的构造背景。沉积序列表明,扬子西缘寒武系沧浪铺组、西王庙组和二道水组主要由砂岩和白云岩等组成,沉积环境为滨岸—潮坪。细—粗砂岩碎屑颗粒为次棱角状—次圆状,分选较差;碎屑组分主要为石英,岩屑几乎全部为燧石,长石含量较少。测试分析结果表明：重矿物分析指示扬子西缘寒武系砂岩重矿物主要由锆石、赤—褐铁矿、电气石、钛铁矿、金红石、磷灰石等组成,重矿物组合指示岩浆岩为其主要母岩;电气石电子探针分析结果表明,物源主要来自于贫锂花岗岩和变砂岩、变泥岩;碎屑锆石测年分析表明物源区母岩主要为983~540 Ma岩浆岩。碎屑锆石年龄对比等综合分析表明,寒武系沉积物部分源自康滇古陆983~708 Ma的岩浆岩和变沉积岩,部分源自冈瓦纳大陆东非造山带663~540 Ma的岩石,物源区岩石经历再旋回产物作用。扬子西缘寒武系的沉积序列、碎屑锆石年龄谱图和碎屑组成等特征综合分析表明,扬子陆块西缘寒武系形成于前陆盆地。     

砂岩副矿物的物源区分析新进展

物源区分析是盆地分析和古地理重建的重要研究方法,有很长的研究历史。随着分析测试技术的进步,物源区分析的方法不断发展,分析精度也不断提高。电子探针、离子探针和LAICPMS等微区原位分析技术的发展,使得基于单矿物地球化学和同位素年代学的物源区分析方法取得了长足的进步,代表了物源区分析的最新发展方向。本文介绍了砂岩中常见副矿物铬尖晶石、金红石的地球化学分析以及锆石UPb年龄、白云母ArAr年龄在物源区分析中的应用。同时,作者认为物源区分析往往是基于碎屑矿物的化学成分或年龄与可能物源区的对比,且只具有统计意义。因此,深入了解区域地质情况和可能的物源区的地质信息,测试一定数量的颗粒进行统计分析,才能得出合理的结果。     

On the scarcity of >3900 Ma detrital zircons in ≥3500 Ma metasediments

Constant volume models for the continental crust require a flux of crustal material back into the mantle (recycling), equal in volume to that of the juvenile igneous suites added to the continental crust throughout time. In growth of crustal volume models, there is not equilibrium between the volume of juvenile crustal additions and any recycling (destruction) of crust. By establishing the proportion of >3900 Ma detrital zircons in early Archaean sediments it might be possible to constrain the relative importance of crustal growth and recycling. Gneiss complexes in western Greenland, northern Labrador and northeastern China contain rare ≥3500 Ma detrital metasediments. In sediments deposited between 3500 and 3600 Ma, ≥3900 Ma zircons have not been detected in a suite of 117 detrital grains. Based on statistical considerations, at the 95% confidence level any ‘missed’ ≥3900 Ma component forms <3% of this suite. Likewise, >3900 Ma detrital grains do not occur amongst 54 detrital grains from (even rarer) 3700–3800 Ma sediments, arguing with 95% certainty that any ‘missed’ ≥3900 Ma component forms <5% of this most ancient suite. If the age spectra of these detrital zircon suites are representative of the complexes in which they reside, then constant volume (recycling=new additions) models require that by 3500 Ma, >97% of >3900 Ma crust was destroyed by recycling. Such an extremely high recycling rate (≈25% of the crust 100 Ma⁻¹) is hard to reconcile with the diversity of initial Nd and Sr isotopic ratios of well preserved early Archaean granitoid suites in the same complexes, because significant average crustal residence times are required to permit the radiogenic isotopic systems to evolve. The most likely interpretation of the detrital zircon record in the Greenland, Labrador and China sediments is that in their provenance areas the volume of continental crust was small at 3900 Ma, and that it grew significantly during the early Archaean. If the measured ≥3500 Ma detrital sediment suites are globally representative, they support growth models for the continental crust in the early Archaean, rather than models involving recycling of a voluminous >3900 Ma sialic crust. Because of its global coverage and the dating of thousands of grains, the age spectra for detrital zircons from 3000–3200 Ma sediments provides a more reliable impression of crustal ages. However, as they were deposited 700–900 Ma after 3900 Ma, the globally small proportion of ≥3900 Ma detrital grains in them (from Jack Hills, Mt. Narryer and Wyoming) can be accommodated in both crustal growth and moderate recycling models.     

Provenance bias between detrital zircons from sandstones and river sands:A quantification approach using 3-D grain shape,composition and age

Preservation bias may significantly impact the application of detrital zircon geochronology in reconstructing Earth surface processes.Here we compare detrital zircons from the actively eroding Murchison River channel in Western Australia with Ordovician fluvial sediments that have drained similar source rocks along the western margin of the West Australian Craton.In addition to standard analysis of detrital zircon age spectra we apply multivariate statistics to test the relation between 3-D grain shape,U-content and U-Pb ages,with the objective to quantify differences between both sample groups and track preservation along the transport pathway of the Murchison River.Our results show that zircon grains in modern river sands display an upstream trend toward larger surface areas,volume equivalent diameters and grain widths,as well as toward higher U-contents and lower apparent grain densities.3-D grain shape,size and age spectra of Murchison River zircons evolve consistently downstream,but even at the river outlet remain distinct from the Ordovician samples,as a less mature representation of source.We interpret Ordovician river zircons to represent a significantly depleted subset from which up to 22% of the zircon population may have been lost compared to the actively transported detrital load.This discrepancy between the characteristics of detrital zircons in modern active rivers and ancient fluvial Ordovician sandstones demonstrates a bias that could be relevant for other source-sink detrital transport systems throughout Earth history.     

渤海钻孔物源示踪和河流沉积物扩散研究:碎屑锆石U-Pb年龄和磷灰石原位地球化学元素双重约束

锆石和磷灰石是河流沉积物中常见的副矿物,由于各自的U-Pb年龄组成和原位地球化学元素组成在不同区域内存在显著差异,是进行河流物源示踪研究的理想矿物。基于此,利用在环渤海湾盆地主要汇入河流已发表的碎屑锆石U-Pb年龄,结合盆地晚第四纪钻孔的近地表碎屑物质的锆石U-Pb年龄,综合Kolmogorov-Smirnov统计方法的多维判别图（MDS）,系统对比分析了辽东湾、渤海中央盆地和莱州湾的物质来源。结果显示,辽东湾的物质主要来自辽河;渤海中央盆地和莱州湾的碎屑物质主要来自黄河。同时,利用激光剥蚀电感耦合等离子质谱仪（LA-ICP-MS）,对黄河下游（n=70）、胶东半岛的威海湾（n=120）、银滩湾（n=60）的现代河流沉积物和海岸砂开展了碎屑磷灰石微区原位（in situ）主微量元素分析。结果发现黄河与威海湾、银滩湾的碎屑物质不存在物源关系,结合该区域碎屑钾长石原位主微量元素的已有分析结果,进一步说明黄河与胶东半岛的海湾内的碎屑物质不存在物源关系。新的研究结果表明,将碎屑锆石、磷灰石原位地球化学分析相结合有助于更精准地判定河流的物源关系。      

石榴石微量元素地球化学及其在沉积物源分析中的应用

石榴石是沉积物中常见的重矿物,其可来源于多种岩石,而且不同类型母岩中石榴石具有多样的地球化学组成,因此碎屑石榴石的地球化学分析在沉积物源研究中应用广泛。通过电子探针分析可以容易地获得单颗粒碎屑石榴石的主量元素地球化学组成,可借此探讨其母岩类型,但也存在一定的局限性,比如中酸性火成岩和部分变沉积岩来源的石榴石通常都具有高Fe、Mn的特征,不易于区分。本文系统地收集了不同岩石类型的石榴石微量元素数据,尝试利用微量元素地球化学的差异性对碎屑石榴石物源分析进行补充。最终得出以下结论:(1)石榴石的稀土元素(REE)组成与钇(Y)元素指标可区分中酸性火成岩和变沉积岩来源的碎屑石榴石;(2)基性岩(橄榄岩、辉石岩)及所对应的变基性岩石(榴辉岩)中石榴石的微量元素地球化学组成相近,但部分橄榄岩来源的石榴石在镨/钬(Pr_N/Ho_N)值和重稀土总量(ΣHREE含量)上与辉石岩和榴辉岩的有显著差别,这一特点可运用于以基性岩母岩为主的碎屑沉积物源研究中;(3)夕卡岩中的石榴石在主量元素地球化学组成上表现为高度一致的高Ca特征,而稀土元素组成具有两种典型的分配模式,岩浆型(指示富铁、氧化环境)与热液型(指示富铝、还原环境)。综上所述,石榴石微量元素地球化学可以有效地运用于沉积物源分析研究中,是其主量元素物源分析方法的重要补充。     

碎石颗粒形状测量与评定的初步研究

颗粒形状是影响碎石料密实特性及力学、渗流特性的因素之一。选取粒径为2～5 mm和5～10 mm的两组灰岩碎石颗粒样本作为研究对象,采用影像测量仪和特制夹具,获取不同旋转角度下的颗粒轮廓影像;使用图形处理软件获得颗粒几何尺寸测值;计算获得各旋转角度下常用颗粒形状评定参数值,运用其平均值进行统计分析,避免了依据单一角度测值评定伴随的人为因素影响。结果表明,灰岩碎石颗粒与标准圆有较大差异, 且粒径大者差异性更明显;两组样本颗粒形状参数均服从偏态分布;长宽比、扁平度和球形度能够更敏感地反映颗粒偏离球形颗粒的程度,而长宽比和球形度便于获取,因而更具优势。     

冰川沉积类型识别与ESR测年样品采样规范

电子自旋共振(ESR)技术是一种确定物质成分和结构的顺磁性质的分析方法,也能够用于沉积物定年。该方法的测量技术和测年的物理机制等还处于发展阶段。冰川作用过程十分复杂,形成各种类型的冰川沉积物,其顺磁信号的归零机制有显著差异,ESR测年的实验方案也有所差异。因此,识别冰碛物类型,采集合适的样品对于ESR测年的准确性十分重要。冰下融出碛和滞碛经过了冰下磨蚀过程,结构致密,细颗粒基质含量高,石英砂中的一些杂质芯的ESR信号能够衰退。许多冰上融出碛,结构疏松,但细颗粒基质含量高,不但经过了搬运过程中的冰下磨蚀过程使ESR信号衰退,又经历了沉积时的冰上阳光直射过程使信号衰退,一些样品的ESR信号能够完全晒退。冰水湖泊和冰水河流沉积的细砂和粉砂来源于冰下研磨的产物,信号会衰退;在搬运沉积过程中又可能被阳光直射,信号进一步衰退。其它类型的冰碛物的ESR信号衰退机制不明,或粒径不适合用ESR方法测年。采集冰碛物ESR年代样品时,最好同时采集信号衰退机制相同的现代冰碛物样品,以便对照,并用于扣除可能的残留信号。     

定量表征河流砂颗粒的形状——在西藏帕隆藏布的尝试

河流搬运体系中,颗粒的形状属性可以反映其来源、搬运、沉积的过程。长期以来沉积学界缺乏相对统一的定量的、多维的颗粒形状参数。本研究以西藏雅鲁藏布江支流帕隆藏布为例,运用图像分析、软件分析等计算机辅助技术,对8个天然河流砂样品中的2276个颗粒开展碎屑成分鉴定和形状参数量化等工作。重点评估了11个颗粒形状参数,通过因子分析等统计方法提取并定义了3个形状信号,包括圆形度因子、规则度因子和平滑度因子。这些参数可用于表征颗粒宏观形状上接近圆形的程度、中观尺度上轮廓规则的程度及轮廓平滑的程度。挖掘天然河流砂的重要形状特征后,进一步探索了这些形状参数与颗粒成分、粒度的关系。本研究成果为解释河流系统中颗粒的形状信号、理解颗粒自身微观属性和外部物理过程等提供重要借鉴。     

北祁连北大河沉积物碎屑组成及物源正演分析

开展现代河流沉积物的物源分析正演研究,对厘清河流搬运分选过程对不同碎屑成分的影响、不同源区对汇区沉积物的贡献机制以及物源定量分析具有重要意义。对北祁连北大河与洪水坝河现代河流沉积物的物源开展了正演研究,结合砾石成分统计、重矿物分析和碎屑锆石U-Pb测年方法,分析了流域内不同物源对现代河流不同碎屑成分贡献的差异性。结果表明,两条河流沉积物的特征显示汇区不同碎屑成分对物源区岩性的反映各有侧重。其中砾石岩性能够反映整个流域的主要岩性;砂粒中的重矿物组合也可以反映源区主要岩性,且对重矿物含量较高的岩体或地层（如铁矿）十分敏感;碎屑锆石年龄谱反映了整个流域物源端元锆石的综合贡献,各年龄峰锆石占比与源区各年龄峰锆石贡献呈较好的正相关关系,其中源区沉积岩碎屑锆石的贡献比较显著。通过河流物源正演分析进一步建立现代河流源汇模式是实现物源量化分析的一个可行方向。     

定量表征河流砂颗粒的形状——在西藏帕隆藏布的尝试

河流搬运体系中,颗粒的形状属性可以反映其来源、搬运、沉积的过程。长期以来沉积学界缺乏相对统一的定量的、多维的颗粒形状参数。本研究以西藏雅鲁藏布江支流帕隆藏布为例,运用图像分析、软件分析等计算机辅助技术,对8个天然河流砂样品中的2276个颗粒开展碎屑成分鉴定和形状参数量化等工作。重点评估了11个颗粒形状参数,通过因子分析等统计方法提取并定义了3个形状信号,包括圆形度因子、规则度因子和平滑度因子。这些参数可用于表征颗粒宏观形状上接近圆形的程度、中观尺度上轮廓规则的程度及轮廓平滑的程度。挖掘天然河流砂的重要形状特征后,进一步探索了这些形状参数与颗粒成分、粒度的关系。本研究成果为解释河流系统中颗粒的形状信号、理解颗粒自身微观属性和外部物理过程等提供重要借鉴。     

珠江口盆地惠州凹陷古近纪多幕裂陷旋回的沉积物源响应

为阐明惠州凹陷古近纪构造活动对物源供给与沉积充填的控制作用,以LA-ICPMS锆石U-Pb定年为主要技术手段,对惠州凹陷古近系不同层段的8件样品中的锆石形态及年代组成进行分析,定量研究其物源供给规律.南部物源,即盆内基岩凸起与东沙隆起,以中生代火成岩为主;北部物源,即盆外华南褶皱带,岩性及年代学复杂.结果显示从下文昌组-上文昌组-恩平组的砂岩样品中,中生代锆石颗粒含量逐渐降低,前寒武-古生代锆石颗粒含量逐渐增加,表明惠州凹陷具有早期南部供源逐渐到晚期北部供源的转换规律.惠州凹陷构造演化分为3幕:裂陷Ⅰ_A、Ⅰ_B和Ⅱ幕.裂陷Ⅰ_A幕,下文昌组沉积时期,凹陷南部断层活动强烈,砂岩中锆石以中生代颗粒为主,沉积相类型以靠近南部断裂带发育的扇三角洲及中深湖相为主;裂陷Ⅰ_B幕,上文昌组沉积时期,北部边界断裂带活动强度增加,中生代锆石颗粒含量为68%,前寒武-古生代颗粒含量为32%,表明该时期南北同时供源,沉积体系以扇三角洲、辫状河三角洲及中深湖相为特色;裂陷Ⅱ幕,恩平组沉积时期,北部断层持续活动,前寒武-古生代锆石颗粒含量为64%,中生代颗粒含量为36%,证实恩平组沉积时期以北部物源为主,发育了浅水辫状河三角洲、滩坝以及薄煤层.      

Quantitative characterization of grain shape: Implications for textural maturity analysis and discrimination between depositional environments

Grain shape plays an important role in textural analysis of sedimentary grains. Textural analysis helps to determine the formation, transportation and deposition processes of sedimentary rocks. However, there is a lack of standardized methodology for quantitative characterization of grain shapes. The utility of fully automated image analysis for grain shape measurement is assessed in this paper. This research aimed to identify the most useful shape parameters for textural characterization of populations of grains and determine the relative importance of the parameters. A key aspect of this study is to determine whether, in a particular sedimentary environment, textural maturity of the samples can be ranked based on their grain shape data. Furthermore, discrimination of sedimentary depositional environments is explored on the basis of grain shape. In this study, 20 loose sediment samples from four known depositional environments (beach, aeolian, glacial and fluvial) were analysed using newly implemented automatic image analysis methods. For each sample, a set of 11 shape parameters were calculated for 200 grains. The data demonstrate a progression in textural maturity in terms of roundness, angularity, irregularity, fractal dimension, convexity, solidity and rectangularity. Furthermore, statistical analysis provides strong support for significant differences between samples grouped by environment and generates a ranking consistent with trends in maturity. Based on novel application of machine learning algorithms, angularity and fractal dimension are found to be the two most important parameters in texturally classifying a grain. The results of this study indicate that textural maturity is readily categorized using automated grain shape parameter analysis. However, it is not possible to absolutely discriminate between different depositional environments on the basis of shape parameters alone. This work opens up the possibility of detailed studies of the relationship between textural maturity and sedimentary environment, which may be more complicated than previously considered.     

Threshold of motion and settling velocities of mollusc shell debris: Influence of faunal composition

Bioclastic particles derived from mollusc shell debris can represent a significant fraction of sandy to gravelly sediments in temperate and cool‐water regions with high carbonate productivity. Their reworking and subsequent transport and deposition by waves and currents is highly dependent on the shape and density of the particles. In this study, the hydrodynamic behaviour of shell debris produced by eight mollusc species is investigated for several grain sizes in terms of settling velocity (measurements in a settling tube) and threshold of motion under unidirectional current (flume experiments using an acoustic profiler). Consistent interspecific differences in settling velocity and critical bed shear stress are found, related to differences in shell density, shell structure imaged by scanning electron microscopy and grain shape. Drag coefficients are proposed for each mollusc species, based on an interpolation of settling velocity data. Depending on the shell species, the critical bed shear stress values obtained for bioclastic particles fall within or slightly below empirical envelopes established for siliciclastic particles, despite very low settling velocity values. The results suggest that settling velocity, often used to describe the entrainment of sediment particles through the equivalent diameter, is not a suitable parameter to predict the initiation of motion of shell debris. The influence of the flat shape of bioclastic particles on the initiation of motion under oscillatory flows and during bedload and saltation transport is yet to be elucidated.     

The role of grain dissolution and diagenetic mineral precipitation in the cycling of metals and phosphorus: A study of a contaminated urban freshwater sediment

This paper describes the detrital mineralogy, early diagenetic reactions and authigenic mineral precipitates for freshwater contaminated sediments deposited in an urban water body (the Salford Quays of the Manchester Ship Canal, Greater Manchester, UK). These sediments contain a mix of natural and anthropogenic detrital grains. Detrital grains are dominated by quartz and clay grains, whilst anthropogenic grains are dominated by metal-rich glass grains, concentrated at a depth of 12–17 cm in the sediment as a result of historical inputs. Sediment porewaters contain significant concentrations of Fe, Mn, Zn and phosphate. Bacterial Fe(III) and Mn(IV) reduction are hypothesised to supply Fe²⁺ and Mn²⁺ to porewaters, with phosphate released from Fe oxide reduction or organic matter oxidation. Petrographic observations indicate that the metal-rich glass grains are undergoing chemical dissolution during early diagenesis, supplying Fe and Zn to porewaters.     

A Study of Pretreatment Methods for Terrigenous Grain-Size Analysis of Marine Sediments

The prerequisite for obtaining variations of terrigenous grain-size of marine sediments is how to effectively remove non-terrigenous matters and preserve terrigenous particles synchronously. Combined with observations under biological microscope and scanning electron microscope, a comparative study of biogenic debris removal effect and terrigenous grain-size analysis under different pretreatment condition was performed on core sediments, which were retrieved in the South China Sea during the MD190 cruise. Our new results showed that the main three biogenic particles, namely, organic matter, carbonate, and opal in marine sediments could be removed effectively by 30% H₂O₂ in a stirring water bath at 60 ℃ for 3 h, 0.5% HCl for 1 h, and 2 mol/L Na₂CO₃ in a stirring water bath at 85 ℃ for 5 h, in turn. Such pretreatments achieved the goals of biogenic debris removal efficiency and relatively well-preserved terrigenous particles. Prior to selecting an appropriate pretreatment method, this study suggested that the actual effects of biogenic detritus on grain-size results of diverse marine sediment samples should be taken into account. If the laboratory data are ensured to be closer to the natural grain-size distribution of terrigenous particles, the removals of all biogenic debris are not always needed, and the less pretreatment processes the better. For example, opal particles have little effect on terrigenous grain-size distribution when their percentage is lower than 2%. Thus, there is no use to remove them from marine sediments before laboratory grain-size analysis of terrigenous particles. Additionally, ultrasonic is not suggested through the whole process of terrigenous grain-size analysis because the strong energy of ultrasonic can lead to the fragmentation of some fragile terrigenous particles.     

砂土颗粒形状的傅里叶描述符自动生成

颗粒形状是影响砂土密实度、力学与渗流等特性的主要因素之一。傅里叶描述法是一种有效表征颗粒形状的数学方 法。基于傅里叶系数与颗粒平均半径定义傅里叶描述符Dn。采用傅里叶描述符D2,D3,D8以及D3与D8对数线性组合的简化 算法可自动生成复杂的砂土颗粒形状。颗粒形状由傅里叶描述符Dn与相位角δn序列共同控制。分析了傅里叶描述符D2、 D3、D8与颗粒几何参数间的相关性,结果表明,长宽比a与D2正相关,圆形度C 与D2,D8负相关,磨圆度R与D8强负相 关,球形度S与D2,D3之间均呈现显著负相关,规则度Re与D8呈负相关性,且在D8=0时与D3之间呈强负相关性。以显微CT 扫描南京粉砂断层序列二值图像中典型颗粒为样本,基于Pearson相关性准则,计算了重构颗粒与实际颗粒轮廓相似性,相 关系数大于0.94,表明算法具有较高的表征精度。算法可用于二维数值模拟中实际砂土颗粒的批量自动生成。