BLOG · 2026-11-11 · 技术分享Technical Notes

从人工铲锤到视觉清渣机器人:阳极组装磷生铁清理工位的工艺与备件变化From Manual Chipping to Vision-Guided Cleaning: Cast-Iron Slag Removal in Anode Assembly

2026 年 10 月 28 日《山西日报》与 11 月 2 日中国铝业网站先后报道,阳极组装车间的磷生铁清渣工位已用三维视觉识别、多自由度机械臂与密闭除尘替代人工铲锤,单厂报告清理效率较人工提升近一倍。本文把工位问题拆成三层:溢铁与板结铁渣从何而来、自动化改造改变了哪些设备与备件、以及熔炼温度优化与常温导电糊组装试验会把需求推向何处。文中同时对比两组公开配方数据的口径差异,并给出可测量的验收指标建议。Shanxi Daily (28 October 2026) and the China Aluminium corporate news site (2 November 2026) both describe an anode-assembly workshop where manual chipping of cast-iron slag has been replaced by 3D vision, a multi-axis robot arm and enclosed dust extraction; the plant reports cleaning efficiency nearly doubled. This note separates three layers: why overflow iron and hardened slag form, which equipment and wear parts the retrofit changes, and where melt-temperature optimisation and room-temperature conductive-paste assembly trials may shift demand. Two published composition ranges are compared to show why they must not be merged into one recipe, and measurable acceptance criteria are proposed.

本栏目前两篇分别讨论了铁—碳压降的成因链条,以及公开采购公告里出现的阳极组装车间改造需求。这一篇把视角收窄到一个具体工位:磷生铁浇注之后的溢铁与板结铁渣清理。2026 年 10 月下旬至 11 月初,中国铝业网站与《山西日报》先后报道了同一类改造——用搭载三维视觉的清渣机器人替代人工铲锤;同期公开的技术文章与行业会议报告,则提供了从熔炼参数到连接工艺替代的两条补充线索。以下把工位问题、设备构成、上游控制与备件影响放在一起梳理,所有数字均标注出处与口径,单厂案例数据不当作行业平均值。

一、清渣工位为什么值得单独看

钢爪与炭块之间靠磷生铁浇注形成磷铁环连接。浇注时铁水容易溢出,在炭块表面及炭碗周边形成板结铁渣,必须在阳极进入电解槽前清除。《山西日报》2026 年 10 月 28 日的报道写明,这一工位以往靠职工手持铲、锤作业,容易出现清理不彻底或磕碰损伤炭块,单班需配置多人,且长期面对高温与粉尘。中国铝业网站 11 月 2 日的企业动态进一步指出,人工清理质量不稳定会损伤炭块、影响阳极导电性能,进而造成电解槽工况波动。

换句话说,清渣不是单纯的外观整理:残留铁渣会改变阳极上槽后的装配与导电状态,磕碰损伤则直接进入废块或影响换极周期。这也解释了为什么改造目标被表述为「统一清理标准」,而不只是「省人」。

二、视觉清渣机器人的技术构成

模块公开报道中的功能工程含义
三维视觉识别视觉扫描自动识别每块阳极的溢铁位置及厚度清理对象从固定轨迹变成按件识别;镜头视窗与光源须在高温、高粉尘下保持可用
轨迹规划与机械臂自主规划清理轨迹,多自由度机械臂清除炭碗、钢爪周边结块磷生铁末端铲具、锤具成为新的易损件;力度控制直接关系炭块是否受损
密闭除尘与废渣收集粉尘即时吸走、废渣集中回收滤材、密封件、输送与收集装置进入日常维护清单
产线配合改造结合现场板链输送产线布局到位精度、夹紧与节拍匹配是视觉定位的前提

报道给出的效果包括:清渣岗位实现无人值守、清理标准统一、避免炭块破损与铁渣残留、稳定阳极导电质量、清理效率较人工提升近一倍、废渣集中回收便于资源化利用。需要强调,这是单一车间改造后自述的结果,不是可外推的通用指标;不同产线的溢铁量、阳极规格与节拍差异很大。

三、把溢铁量压在上游:熔炼与浇注参数

自动化解决的是「清得干净」,更经济的方向是「少产生」。河北网络广播电视台 2026 年 10 月 2 日刊出的技术文章把铁水质量归结为原材料成分、熔炼温度与设备状态三者共同作用:传统高硅、高磷配方熔炼温度通常达 1400~1450℃,部分情况超过 1500℃;改用高碳、低硅、低磷优化配方后可降至 1250~1350℃,下降约 100~150℃,相关生产试验中试验阳极的铁—炭压降平均降低约 30mV,并减轻炉衬高温侵蚀及铁环压脱负荷。文章建议的措施包括按批次炉料化验动态配料、炉前快速光谱分析后增碳或脱硫调质、出铁前连续测温,以及用自动化温控浇铸设备保持恒定的浇注速度与连续流速。

配方数字要谨慎使用。该文列出的公开区间为 C 2.5~3.5%、Si 2.0~3.0%、Mn 0.6~0.9%、P 0.8~1.6%、S 低于 1.5%;而一份较早的中频感应电炉浇注施工工法资料给出 P 0.5~1.4%、S 低于 0.2%,并要求钢爪与炭碗预热不低于 100℃、钢爪与炭碗最小间隙不小于 10mm、浇包浇注前必须烘干、磷铁环高出炭碗上表面 3~5mm。两组数字对应的年代、槽型与统计口径都不同,且「熔炼温度」与「出炉/浇注温度」不是同一个量,不能拼接成一张工艺卡。它们只说明一件事:流动性、过热度、组对间隙与干燥状态共同决定溢铁与溅渣的多少,实际控制值应回到本厂化验单与工艺卡。

四、再往远看:连接工艺改变时,工位需求会重新分配

2026 年 10 月 14 日铝业周的一场辅料市场研讨会上,一份关于导电糊工业试验的报告给出另一条路径:用阳极导电糊常温组装钢板阳极,工业试验中阳极压降较现行磷生铁浇注钢爪阳极降低 70~150mV,吨铝节电约 100~240 度,电流效率提高约 0.2%~0.5%;报告同时提到高导糊用量较少、可减少残极污染、钢板更易清理,目前处于扩大试验阶段。同场协会报告披露,截至 2026 年 9 月预焙阳极总产能 3265 万吨,石墨化阴极产量占比升至 77.6%、电解槽采用比例达 83%。

这组数据的意义不在于预言某条产线会立刻切换,而在于说明存量基数很大:即便常温组装长期成立,磷生铁浇注及其配套清理、压脱、抛丸、除尘能力在相当长时间内仍要维护。而若连接形式变化,需求会从「清除板结铁渣」转向钢爪或钢板的表面清理、涂层与装配精度,以及残极处理环节的重新配置。上述试验数据来自单方会议报告,尚未形成行业普及结论。

五、对备件与检修窗口的实际含义

  • 机器人侧:视觉视窗与光源、末端铲锤工具、密闭除尘滤材与密封、废渣收集输送部件,属于高温高粉尘工况下的新增维护项;标定周期与备件供货周期宜在改造验收时一并约定。
  • 传统链上:抛丸器叶片与护板、压脱模具、液压密封、除尘滤袋仍是残极与磷铁环处理工位的高频更换件,自动化改造不会取消这些工位,只是改变其入口物料状态。
  • 熔炼与浇注侧:中频炉炉衬捣打与烘炉烧结质量、连续测温元件、炉前光谱取样、浇包烘烤与耐火材料,是把质量前移时必须同步保障的项目;技术文章特别提示预防性巡检可避免磁场分布不均造成局部过热或炉衬破损。
  • 验收口径:建议把「清理标准」转成可测量指标(残铁残留量、炭块磕碰率、压降抽测),再据此判断是补单机、改工位,还是调整工艺纪律。

以上为公开资料整理与一般性工程观察,不构成针对具体产线的技术建议;文中涉及的公开报道、会议数据与工艺参数均不属于本公司的业务、客户或业绩,具体以原发布方页面和现场工艺文件为准。

The previous two notes in this column examined the cause chain behind iron-to-carbon voltage drop and the retrofit demand visible in public procurement notices for anode assembly workshops. This one narrows the view to a single station: removing spilled cast iron and hardened slag after the phosphorus-iron pour. In late October and early November 2026, the China Aluminium corporate news site and Shanxi Daily both reported the same type of retrofit, replacing manual chipping with a vision-guided cleaning robot. Technical articles and an industry conference report published in the same window add two further threads, one upstream in melt control and one further out in alternative joining methods. Every figure below is labelled with its source and scope, and results from a single workshop are not treated as industry averages.

1. Why this station deserves its own analysis

The steel claw and the carbon block are joined by a cast-iron ring formed when molten phosphorus iron is poured into the gap. The melt readily overflows and forms hardened slag on the block surface and around the carbon bowls, which must be removed before the anode enters the reduction cell. Shanxi Daily, on 28 October 2026, described the station as one where workers used hand-held shovels and hammers, with incomplete cleaning or impact damage to the block as recurring problems, several operators needed per shift, and sustained exposure to heat and dust. A China Aluminium corporate news item on 2 November added that unstable manual cleaning can damage blocks, impair anode conductivity and ultimately disturb cell operating conditions.

Cleaning is therefore not cosmetic. Residual iron changes how the anode seats and conducts after it is set, while impact damage feeds directly into scrap blocks or shorter change cycles. That is also why the stated goal of the retrofit is a consistent cleaning standard rather than headcount reduction alone.

2. What the vision-guided cleaning cell actually contains

ModuleFunction reported publiclyEngineering implication
3D visionScans each anode to identify the position and thickness of overflow ironCleaning shifts from a fixed path to per-piece recognition; camera windows and lighting must survive heat and heavy dust
Path planning and robot armPlans its own trajectory; a multi-axis arm removes lumped cast iron around bowls and clawsEnd-effect chisels and hammers become new wear parts; force control determines whether the block is damaged
Enclosed dust extraction and slag collectionDust is captured as it is generated; slag is collected centrallyFilter media, seals, conveyors and collection units enter the routine maintenance list
Line integrationThe retrofit was designed around the existing chain-conveyor layoutPositioning accuracy, clamping and cycle-time matching are preconditions for reliable visual location

The reported outcomes include unattended operation, a uniform cleaning standard, fewer damaged blocks and less residual iron, more stable anode conductivity, cleaning efficiency nearly double that of manual work, and centralised slag recovery for recycling. These are self-reported results from one workshop, not a transferable benchmark: overflow volume, anode format and line cadence vary considerably between plants.

3. Reducing overflow upstream: melting and pouring parameters

Automation addresses cleaning quality; the more economical objective is to generate less slag in the first place. A technical article published by Hebei Network Broadcasting on 2 October 2026 attributes melt quality to three factors acting together: charge composition, melting temperature and furnace condition. Conventional high-silicon, high-phosphorus practice typically melts at 1,400-1,450 degrees Celsius and sometimes above 1,500 degrees; with an optimised high-carbon, low-silicon, low-phosphorus composition the range drops to 1,250-1,350 degrees, roughly 100-150 degrees lower. In the production trials cited, the average iron-to-carbon voltage drop of test anodes fell by about 30 mV, with less high-temperature attack on the lining and lower load on ring stripping. Recommended measures include dynamic charge calculation from batch-by-batch analysis, rapid spectrometric checks at the furnace with carbon or sulphur correction, continuous temperature measurement before tapping, and automated temperature-controlled pouring that holds a constant pour speed and uninterrupted flow.

Composition figures need particular care. That article lists C 2.5-3.5%, Si 2.0-3.0%, Mn 0.6-0.9%, P 0.8-1.6% and S below 1.5%. An earlier construction-method paper for medium-frequency induction furnace pouring gives P 0.5-1.4% and S below 0.2%, and calls for preheating claws and bowls to at least 100 degrees Celsius, a minimum claw-to-bowl gap of 10 mm, ladle drying before pouring, and a cast-iron ring standing 3-5 mm above the bowl surface. The two sets come from different eras, cell sizes and definitions, and melting temperature is not the same quantity as tapping or pouring temperature; they cannot be merged into one process card. What they do establish is that fluidity, superheat, assembly gap and dryness jointly determine how much overflow and spatter is produced. Actual setpoints belong on each plant's own laboratory sheets and process cards.

4. Further out: alternative joining would redistribute the work

At an ancillary-materials seminar held on 14 October 2026 during China Aluminium Week, a report on industrial trials of conductive paste set out a different route. Room-temperature assembly of steel-plate anodes with an anode conductive paste reduced anode voltage drop by 70-150 mV compared with conventional cast-iron-poured claw anodes, saved roughly 100-240 kWh per tonne of aluminium and raised current efficiency by about 0.2-0.5%. The same report noted that a high-conductivity paste uses less material, reduces butt contamination and leaves steel plates easier to clean, with expanded trials still under way. An association report at the same event put total prebaked anode capacity at 32.65 million tonnes as of September 2026, with graphitised cathodes at 77.6% of cathode output and installed in 83% of cells.

The point is not that any line will switch immediately, but that the installed base is large. Even if room-temperature assembly proves durable, cast-iron pouring and its cleaning, stripping, shot-blasting and dust-extraction equipment will need maintaining for a long period. If the joint does change, demand moves away from removing hardened slag towards surface preparation of claws or steel plates, coating and assembly accuracy, and a reconfigured butt-handling stage. The trial figures come from a single conference report and are not yet an industry-wide conclusion.

5. What this means for wear parts and shutdown windows

  • On the robot side: vision windows and lighting, end-of-arm chipping tools, extraction filter media and seals, and slag collection and conveying components are new maintenance items in a hot, dusty environment. Calibration intervals and spare-part lead times should be agreed as part of retrofit acceptance.
  • On the conventional chain: blast-wheel blades and liners, stripping dies, hydraulic seals and dust bags remain the high-frequency replacement parts of butt and cast-iron-ring handling. Automation does not remove these stations; it changes the condition of the material arriving at them.
  • On melting and pouring: lining ramming and sintering quality, continuous temperature sensors, spectrometric sampling, ladle preheating and refractories must be secured if quality is to be moved upstream. The technical article specifically warns that preventive inspection avoids uneven magnetic fields, local overheating and lining failure.
  • Acceptance criteria: translate a qualitative cleaning standard into measurable indicators such as residual iron, block impact-damage rate and sampled voltage drop, then decide whether to add a machine, rework a station or tighten process discipline.

This is a compilation of public material and general engineering observation, not technical advice for a specific line. The reports, conference data and process parameters cited here are not the business, clients or track record of our company; the original publisher pages and each site's own process documents remain authoritative.

公开来源Public sources

来源信息以原发布方页面为准;本文仅作整理与解读,不代表任何招标人或发布方立场。Please refer to the original publisher for authoritative details. This article is an independent summary and does not represent any tenderer or publisher.