晶圆载具换料，最容易出事的地方不在力学性能上，而在两个肉眼看不见的数字上。
去年冬天，一家做半导体载具的客户寄来两个花篮。
件用无尘袋装着，袋口贴了封条，标签上写着一行编号。同一天到的还有一份单页检测报告，纸上是一列有机物和离子的名字。
电话里他问得很直接："料是我自己换的，可这不是料的事吧？换之前一直好好的。"
这正是晶圆载具换料的典型开场：外观合格、尺寸过关、装机也顺，问题出在没人盯着的那两项上。
我先问了他三句话。
"换的是基材，还是连助剂体系一起换了？"
"表面电阻是拿粒子压片测的，还是件洗干净、烘完再测的？"
"烘烤验证跑的是多少度、多少小时？"
他停了一下，说头一句答不上来，后两句没做过。
把这件事拉成一条时间线，会更清楚。
起点是换料后的第一批件：外观合格、槽距过检、装配顺畅，成本还比原来低了一截。
潜伏期在第二个月。光刻区的暗场检查开始偶发地报微粒，量很小，按偶发处理，没人往载具上想。
爆发在第三个月：一批件在客户端被判不合格，整批退回重洗，线停了小半天。
结算是追溯。切片做表面成分分析以后，指向的是迁移到表面的低分子组分，跟力学一个字都不沾。
补一个换算：槽距公差 ±0.05 mm，一只 25 槽的花篮从头排到尾，偏差能累积到 1.25 mm 量级——晶圆不是放不进，是会偏摆。
偏摆就会蹭到槽壁，蹭下来的东西落在晶圆上，就是脱屑。
两分钟看懂：脱屑到底是怎么发生的
材料里本来就有两类"会走动"的东西：一是低分子的助剂，二是断掉的分子链碎片。
它们平时待在树脂里，但加工和使用都要给它们能量——温度一高、溶剂一泡，它们就往表面挪。
挪到表面之后，遇到烘箱那一段高温，它们会一起交出去，粘住空气里飘着的微粒。
一句大白话：脱屑不是件掉下来的碎屑，是材料自己"出汗"之后粘住的灰。
为什么载具场景把这件事放大？因为普通件上它只是一层白雾，载具上它是一整片晶圆的污染源。
两者的后果差好几个数量级，处置成本也一样。
一、换料前，六维里至少四样要落到数字
载具这一件，工况要用温度、介质、洁净、静电、寿命、尺寸六个维度来问。
温度上，它不跟高温工艺走，但要反复进烘箱，烘干段常在 80–120℃；湿法清洗线的药液常压在 60–80℃。
这两段是交替来的，不是一次性通过。一个班次里冷热交替几轮，一年下来就是几万次。
介质上，SC-1 是氨水配双氧水，SC-2 是盐酸配双氧水，后面还可能过稀氢氟酸和超纯水。
酸碱和氧化剂轮着来，这跟一般工业件"偶尔擦一擦"完全是两回事。
洁净是最特殊的一维：它服务的是一片价值高出它几个数量级的晶圆，所以它自己不能往外吐东西。
洁净室按 ISO Class 分档，光刻区与周边常在 Class 3 到 Class 5 之间，越高等级对材料析出的容忍度越低。
静电耗散级常按表面电阻 10⁶–10⁹ Ω 划。低于这个区间偏导电，高于它就变成绝缘体，表面把飘浮微粒吸住不放。
寿命一般按清洗次数定，行业里 500 次是个常见口径。按一周清洗 3 次算，一年约 150 次，500 次大约落在三年多。
尺寸上，槽距公差常见 ±0.05 mm，定位孔和卡槽按丝级管控，而且要在吸湿之后仍然守得住。
六维里温度、介质、洁净、静电这四样先问出数字，再谈换哪条料；四样都没有，换料就是拿一批货去赌。
二、三条路线摆开，不做谁更好的判断
换料不是找"更强的那个"，是把三条路线的代价摆清楚，看哪条跟你的清洗线、洁净等级和模具兜得住。
| 路线 | 体系特征 | 静电怎么给 | 清洗与尺寸 | 常从哪条换过来 |
|---|
| 玻纤增强 PA66 | 刚性足、成本可控 | 多为表面涂覆或后加 | 酸碱交替下界面先露头 | 金属件或通用工程塑料 |
| 矿物填充 PA6 | 各向异性小、收缩稳 | 需本体导电体系配合 | 尺寸波动小、耐清洗一般 | 玻纤增强件 |
| 长碳链 PA12 低析出 | 吸水低、耐化学好 | 本体永久型更易做 | 反复清洗下保持更好 | 通用 PA66 |
三条没有高下，只有匹配。玻纤增强路线给你刚性，但酸碱交替的环境会先去啃纤维和树脂的界面。
矿物填充路线把各向异性压下来，长条件的翘曲好控，代价是刚性让出去一截。
长碳链路线把吸水压到很低，尺寸和清洗都占便宜，代价是单价和加工窗口。
为什么长碳链路线在清洗线上更耐用？酰胺基密度低，亲水基团少，药液能咬住的位置就少。
而玻纤增强体系的失效点往往不在树脂上，在界面上：纤维和树脂靠处理剂连接，药液一轮一轮泡，先把这层连接啃松。
啃松之后纤维就和树脂脱开，纤头露出来，掉屑和不平整都从这里开始。
先写清"原来用的是什么、为什么要换"，路线自己就收窄了。
三、换料要重验的那张表
下面这张表把上面的约束落成可核对的动作。门限是方向性建议，不是验收标准，实际数值要由你的件和你的清洗线定。
| 指标 | 方向性门限 | 验证方法 / 标准 | 换料后常见失效 | 通行解法 | 对应助剂体系 |
|---|
| 表面电阻 | 10⁶–10⁹ Ω，静电耗散级 | 表面电阻仪，参考 IEC 61340 系列 | 电阻漂进绝缘区，吸微粒 | 换本体导电通路 | 永久型抗静电体系 |
| 总析出量 | 低析出体系常见口径 ≤5 μg/g | 热脱附气质联用 | 烘烤后表面出雾、粘微粒 | 高分子量助剂 + 成型后清洗 | 低迁移润滑体系 |
| 清洗液耐受 | 500 次后拉伸保持七成以上 | SC-1 / SC-2 循环浸泡 + ISO 527 | 露纤、掉屑 | 长碳链基材 + 界面处理 | 界面处理剂 |
| 清洗后表面粗糙度 | Ra 变化受控 | 粗糙度仪 / 白光干涉 | 纤头刮伤晶圆边缘 | 纤维表面处理 | 界面处理剂 |
| 吸湿后尺寸 | 槽距与定位孔仍在公差内 | 调湿后三坐标 | 长条件两端翘、装配卡滞 | 低吸水基材 + 矿物填充 | — |
| 擦拭后静电 | 擦拭 5000 次后漂移半个数量级内 | 擦拭后复测表面电阻 | 越擦越绝缘 | 通路做在本体里，不做在表面 | — |
| 离子残留 | 超纯水冲洗后无可见水渍 | 超纯水冲洗 + 洁净烘干 | 水渍留下离子残留 | 清洗与干燥一并定 | — |
这张表怎么用：先看前两行。析出和表面电阻过不去，后面几行测出来的数都没有意义。
因为这两项失效的性质是"件污染了别人的产品"，比件自己坏掉要重。
第三、四行是配套项。清洗耐受和表面粗糙度决定载具能用多久，也决定它会不会去刮晶圆边缘。
第五行是精密件的老问题。吸湿后的尺寸漂移测不出来，往往是因为量测条件没跟材料一起定。
最后两行容易被当成"检测部门的事"跳过，但它们是判定依据本身。
也别一上来就追最高档的材料。低析出专用体系的料号成本，通常是通用牌号的 2–3 倍，先确认你的件真的需要，再花这笔钱。
引用标准时把版本年份一并写进规格书，量测方法不一致，两家的数据就没法放在一起比。
四、换料以后最常见的四种失效
失效一：脱屑。多数人的第一反应是"玻纤加多了，换个低玻纤的"。
但换料后的第一批并不掉屑，是第三个月才出来的——说明问题不在含量，在助剂慢慢往表面走。
这一条要从助剂侧归因：脱模和润滑体系如果用了低分子、易迁移的型号，它会先往表面跑，烘烤再一催，就成了洁净室里那层看不见的雾。
把润滑体系换成高分子量或反应型，脱屑的量级会跟着下来。
失效二：表面电阻越擦越不对。根因通常是抗静电做成了表面涂层或迁移型，擦拭和清洗把通路一起洗掉了。
这种情况在开发阶段测得很漂亮，一量产就掉——因为开发阶段没人擦它五千次。
失效三：同一批件烘烤以后雾得深浅不一。这不是"料不稳定"，常见是助剂分散不均匀，局部浓度高的地方先析出。
同一批里有的件好有的件出雾，方向就指向混料与母粒化，不指向基材。
失效四：长条形的件两端翘。根因在玻纤取向叠加吸湿，不在刚性不足。
想验证这条判断很简单：把件按调湿前后各量一次，如果翘的量在吸湿之后明显放大，就不用去改料了。
这里有一句我想单独说：很多人以为载具的洁净度是洗出来的，其实有一半是配方里选出来的。
清洗只能洗掉表面的存量，洗不掉材料继续往外迁移的趋势。
五、加工与验证，顺序不能换
栽在载具上的客户，多数不是没做验证，是顺序颠倒了。
头一件事是干燥。吸水料含水超标，进料筒就被剪链，表现是件发脆、表面出银纹，而且显形很晚。
我们经手的换料投诉里，有一类特别典型：同一批料、同一个模具，这一模好下一模脆，配方一个字没改，是干燥。
含水率这一项在南方梅雨季最难守。拆包后在车间放几个小时，水分就能回升一截，干燥做得再好，周转敞口也是白做。
我们的做法是上机前用仪器或露点数据确认，不凭手感。
第二件事是模温。模温低，玻纤会被冻在表面形成浮纤，表面粗糙度直接就上去了，后面清洗再一磨，纤头就露出来。
模温这一项对洁净件的意义比一般件更大：一般件浮纤是外观问题，载具上它就是掉屑的起点。
第三件事是成型后的清洗与烘干。这两道工序的参数要跟材料一起定，不能照抄原来的档位。
第四件事才是包装。包装材料本身也可能析出，要用洁净级包装并做验证。
验证顺序建议这样排：
1. 材料级：含水率、干态与湿态强度保持率
2. 表面级：表面电阻、总析出量
3. 工艺窗口：不同模温、不同保压打出来的件做对比
4. 件级：调湿后槽距与定位孔复测、清洗轮次后的粗糙度
5. 客户端：整机装配与产线暗场检查
前一项不通过就往下走，后面测出来的数据没有解释意义。
为什么顺序不能换？因为表面电阻和析出都受成型表面状态影响，表面没锁住就去调清洗轮次，调出来的结论只对那一批有效，放量又漂。
六、边界：这几种情况，先别换
这一段可能比前面几段更值钱，因为它帮你在开工前止损。
其一，检测能力还没到位的产线。析出和表面电阻都要专门设备，如果现在只有称重加目检，换了也证明不了好坏，先把手段补上。
其二，长期泡强氧化性药液、清洗轮次远超 500 次的位置。这一类要看含氟材料或专门的高纯体系，普通改性尼龙撑不到那个轮次。
其三，需要局部透明的部位。观察窗这类位置建议走分体设计，用别的材料单做，不要为了一个窗口把整件料拉高。
其四，槽距公差要压到 ±0.01 mm 级、且要求塑料一体成型的件。这个量级已经超出注塑加吸湿能守的范围，该回到金属或做结构分件。
把这四条写在前面，不是劝退，是省时间——样品顺、批量卡、整案回退，学费比一开始不换高得多。
七、换料风险清单
| 环节 | 要动什么 | 容易漏的点 |
|---|
| 模具 | 收缩率随填充体系变，长条件两端要复算 | 只换料不核收缩，槽距漂 |
| 干燥 | 按实测含水率定窗口，除湿干燥机必备 | 热风干燥对吸水料基本无效 |
| 调湿 | 强制调湿并称重判定，再复测尺寸 | 按平均壁厚估时间，厚壁没吸透 |
| 料温与模温 | 长碳链料与玻纤料窗口不同，要联合调 | 照抄原件的温度档位 |
| 保压与脱模 | 脱模体系换成低迁移型号 | 沿用原来的脱模剂，它成了析出源 |
| 清洗与烘干 | 药液轮次、烘干温度与时间重定 | 沿用旧参数，水渍与离子残留超 |
| 包装 | 换洁净级包装并做验证 | 包装材料本身就是污染源 |
| 验证顺序 | 材料→表面→工艺→件级→客户端 | 前一项未过就往下走 |
八、打样试模排程
给载具换料排试模，通常分三轮，轮次之间不跳步。
第一轮·小样比对：用你的原模具打 3–5 模，只验含水率、外观和短射件的位置。这一轮不追性能，先把"料能不能填进去"确认掉。
留样两件，标注批号与干燥参数，至少留到第二轮结束。
第二轮·表面与尺寸：固定料，变模温与保压，打两组对比件。验表面电阻、总析出量、调湿后的槽距与定位孔。
这一轮基本决定量产参数。留样按批次封存，至少留到量产稳定后三个月。
第三轮·清洗与客户端：按 500 次的口径跑清洗循环，中途复测粗糙度与强度保持率，然后送客户端走装配和暗场检查。
这一轮过了，才建议放量。留样封存覆盖首批量产，方便追因。
三轮的时间账要提前算：第一轮到第二轮通常两周，第三轮看你的清洗线排期，往往要一个月以上。
想砍轮次可以，但砍掉的那一轮，通常会以"批量退回重洗"的形式补回来。
给技术员向上汇报用的一页纸，可以这样收：
| 场景 | 推荐路线 | 关键指标 | 验证标准 | 需先确认的条件 |
|---|
| 花篮、片盒 | 低析出体系 + 本体静电通路 | 总析出量、表面电阻 | 热脱附 + 表面电阻仪 | 洁净等级与烘烤温度 |
| 频繁清洗的内件 | 长碳链低析出体系 | 清洗轮次后保持率 | 循环浸泡 + ISO 527 | 药液种类与轮次 |
| 高精度槽距件 | 矿物填充体系 | 调湿后槽距公差 | 调湿后三坐标 | 量测条件与基准 |
| 承力框架 | 玻纤增强体系 | 力学保持率 | ISO 527 | 静电是否另做处理 |
九、几个被反复问到的问题
问：换料以后要不要重新走洁净认证？
答：要。认证认的是具体牌号，连色母和脱模剂都要在同一档上，换了料就重新报一遍，这条没有捷径。
问：表面电阻能不能靠后处理涂层解决？
答：可以，但清洗和擦拭会把涂层一起带走。轮次多的件，把通路做在本体里更稳。
问：批次一致性怎么做到？
答：载具客户普遍要求关键牌号批批留样，表面电阻、析出和尺寸随批出报告，任何一项漂移就整批挂起。做的是数据链，不是一张价目表。
问：开头那三句话为什么能定方向？
答：头一句分清了换的是基材还是整套体系，第二句把量测条件问出来，第三句确认验证到底做没做。三句答得齐，问题通常已经缩小到一两项上。
配方里的助剂体系按件的工况配——常规助剂常备现货，特殊型号按需配套；你报工况和牌号，料和助剂一次配齐。
关于我们，四句话——选料这件事，越早问越省事。
一句是做改性尼龙的，一句是料和助剂一起配的，一句是先问工况再谈价，一句是敢说哪个件我们不做。
这类件的换料与试模，可以一起聊。
In wafer carrier material replacement, the most prone to problems is not in the mechanical properties, but in two numbers that cannot be seen with the naked eye.
Last winter, a client who makes semiconductor carriers sent two flower baskets.
The item was packed in a dust-free bag, with a seal on the bag opening and a row of numbers written on the label. On the same day, a single-page test report also arrived, with a list of names of organic substances and ions on the paper.
He asked very directly over the phone: 'I changed the parts myself, but this isn't really about the parts, right? Everything was fine before I changed them.'
This is exactly the typical start of wafer carrier material replacement: the appearance is qualified, the dimensions are up to standard, and installation goes smoothly, but the problem lies in those two items that no one is paying attention to.
I first asked him three questions.
Did you change the substrate, or did you change the additive system along with it?
Is the surface resistance measured by pressing the particles into a tablet, or is it measured after the piece has been washed and dried?
What temperature and how many hours is the baking verification run at?
He paused and said he couldn't answer the first sentence and hadn't done the last two sentences.
If you lay this matter out as a timeline, it will be clearer.
The starting point is the first batch of parts after the material change: appearance is qualified, groove pitch passes inspection, assembly is smooth, and the cost is even lower than before.
The incubation period is in the second month. The dark field inspection in the lithography area began to occasionally report particles, the amount was very small, handled as occasional occurrences, and no one thought about it on the carrier.
Outbreak in the third month: A batch of items was deemed unqualified at the client side, the entire batch was sent back for re-washing, and the line was stopped for almost half a day.
Settlement is retrospective. After the slices are used for surface composition analysis, they point to the low-molecular components that have migrated to the surface, which has nothing to do with mechanics at all.
Add a conversion: with a slot pitch tolerance of ±0.05 mm, a 25-slot basket arranged from start to finish can accumulate a deviation up to the 1.25 mm level — the wafer won't be unable to fit, but it will be misaligned.
If it tilts, it will scrape against the groove wall, and the material that rubs off will fall onto the wafer, causing chipping.
Understand in Two Minutes: How Flaking Actually Happens
There are originally two types of 'movable' substances in the material: one is low-molecular additives, and the other is broken molecular chain fragments.
They usually stay in the resin, but processing and use give them energy—once the temperature rises or they soak in a solvent, they move toward the surface.
After moving to the surface, when they encounter the high temperatures in the oven, they will be released together, sticking to the particles floating in the air.
A plain statement: flaking is not some fallen debris; it is the dust that sticks after the material 'sweats' on its own.
Why does the vehicle scene amplify this matter? Because on ordinary parts it's just a layer of white mist, while on vehicles it's a whole wafer of pollution sources.
The consequences of the two differ by several orders of magnitude, and the disposal costs are the same.
1. Before changing materials, at least four items in the six dimensions must be converted to numbers.
For this vehicle, the operating conditions should be asked in six dimensions: temperature, medium, cleanliness, static electricity, lifespan, and size.
In terms of temperature, it does not follow high-temperature processes, but it needs to repeatedly enter the oven, with the drying stage often at 80–120°C; the chemical solution in the wet cleaning line is usually maintained at 60–80°C.
These two phases alternate; they do not happen all at once. In one shift, cold and hot cycles alternate several times, which adds up to tens of thousands of times over a year.
Regarding the media, SC-1 is ammonia mixed with hydrogen peroxide, SC-2 is hydrochloric acid mixed with hydrogen peroxide, and later it may be followed by dilute hydrofluoric acid and ultrapure water.
Acids, bases, and oxidizers take turns coming, which is completely different from the occasional 'wipe' of ordinary industrial parts.
Cleanliness is the most special one-dimensional aspect: it serves a wafer whose value is several orders of magnitude higher than itself, so it cannot expel anything outward.
Cleanrooms are classified according to ISO Class. The photolithography area and its surroundings are usually between Class 3 and Class 5. The higher the class, the lower the tolerance for material outgassing.
Static dissipation grade is usually determined by surface resistance of 10⁶–10⁹ Ω. Below this range, it is slightly conductive; above it, it becomes an insulator, and the surface holds floating particles firmly.
Lifespan is generally determined by the number of washes, and in the industry, 500 times is a common standard. Assuming washing three times a week, that's about 150 times a year, so 500 times would be a little over three years.
In terms of dimensions, the slot pitch tolerance is commonly ±0.05 mm, and the positioning holes and slots are controlled according to thread standards, while still maintaining accuracy after moisture absorption.
In the six dimensions, first ask for the numbers for temperature, medium, cleanliness, and static electricity before discussing which material to change; if none of the four are available, changing materials is just taking a batch of goods to gamble.
2. Lay out three routes without making a judgment about which is better
Changing materials is not about finding the 'stronger one'; it's about clearly laying out the costs of the three routes and seeing which one aligns with your cleaning line, cleanliness level, and mold compatibility.
| Route | System characteristics | How to give a static shock | Cleaning and Size | Which one do you usually switch from? |
|---|
| Glass Fiber Reinforced PA66 | Rigid foot, controllable cost | Mostly surface coating or added afterward | Under alternating acid and alkali, the interface is exposed first. | Metal parts or general-purpose engineering plastics |
| Mineral-filled PA6 | Low anisotropy, stable contraction | Requires coordination with the intrinsic conductive system | Small size fluctuations, generally washable | Glass fiber reinforced component |
| Long carbon chain PA12 with low precipitation | Low water absorption, good chemical resistance | The permanent type of the main body is easier to make | Repeated washing keeps it better | General Purpose PA66 |
There is no superiority among the three; there is only suitability. The fiberglass-reinforced route gives you rigidity, but in alternating acidic and alkaline environments, it will first attack the interface between the fibers and the resin.
The mineral filling route suppresses anisotropy, making it easier to control warping over long conditions, at the cost of some rigidity.
The long carbon chain route reduces water absorption to a very low level, benefiting size and cleaning, at the cost of unit price and processing window.
Why is the long carbon chain route more durable on the cleaning line? The amide group density is low, there are few hydrophilic groups, so there are fewer positions where the chemical solution can grip.
The failure point of the glass fiber reinforced system is often not in the resin, but at the interface: the fibers and resin are connected by a treatment agent, and repeated soaking in the solution gradually loosens this connection.
After biting the pine, the fibers separate from the resin, the fiber tips are exposed, and flaking and unevenness all start from here.
First clearly write down what was originally used and why you want to change it, and the path will naturally narrow on its own.
3. The form that needs to be re-checked when changing materials
The table below translates the above constraints into verifiable actions. The thresholds are directional suggestions, not acceptance criteria; the actual values need to be determined by your parts and your cleaning line.
| Indicator | Directional threshold | Verification Method / Standard | Common failures after material change | Common solution | Corresponding auxiliary agent system |
|---|
| Surface resistance | 10⁶–10⁹ Ω, static dissipation grade | Surface resistance meter, refer to IEC 61340 series | Resistor drifts into the insulation area, attracts particles | Change the main conductive path | Permanent antistatic system |
| Total precipitation amount | Typical caliber of low precipitation system ≤5 μg/g | Thermal desorption gas chromatography-mass spectrometry | Surface fogging and sticky particles after baking | High molecular weight additives Cleaning after molding | low-migration lubrication system |
| Cleaning solution tolerance | After 500 times, stretch and maintain more than 70% | SC-1 / SC-2 cyclic soaking ISO 527 | Exposed fibers, shedding | Long carbon chain substrate Interface treatment | Interface processing agent |
| Surface roughness after cleaning | Ra variation controlled | Roughness Tester / White Light Interference | Fiberglass tip scratch on the wafer edge | Fiber surface treatment | Interface processing agent |
| Dimensions after moisture absorption | The slot pitch and positioning hole are still within tolerance | Coordinate measuring machine after humidity adjustment | Both ends of the long condition are bent, causing assembly jamming | Low water-absorption substrate Mineral filler | — |
| Static electricity after wiping | Drifts within half an order of magnitude after 5000 wipes | Retest surface resistance after wiping | The more you rub, the more insulated it becomes | The pathways are made inside the body, not on the surface. | — |
| Ion residue | No visible water spots after rinsing with ultrapure water | Rinse with ultrapure water Clean and dry | Water stains leave ionic residues | Washing and drying are determined together | — |
How to use this table: first look at the first two rows. If the extraction and surface resistance don't pass, the numbers measured in the following rows are meaningless.
Because the nature of these two failures is 'contaminating someone else's product,' which is more serious than the item itself being defective.
The third and fourth lines are the supporting items. Cleaning resistance and surface roughness determine how long the carrier can last and whether it will scrape the wafer edges.
The fifth line is the old problem with precision parts. The dimensional drift after moisture absorption cannot be measured, often because the measurement conditions were not determined together with the material.
The last two lines are easily dismissed as 'the concern of the inspection department,' but they are the basis for judgment itself.
Don't rush to go for the highest-grade materials right away. The part numbers for low-precipitation specialized systems usually cost 2–3 times that of general-purpose grades, so confirm that your parts really need it before spending that money.
When quoting standards, include the version year in the specification. If measurement methods are inconsistent, the data from the two companies cannot be compared together.
4. The Four Most Common Failures After Material Change
Failure 1: Flaking. Most people's first reaction is 'there's too much fiberglass, switch to one with less fiberglass.'
But the first batch after the material change did not shed flakes; it only appeared in the third month — indicating that the problem is not in the content, but in the additives slowly migrating to the surface.
This point should be attributed to the additives side: if the release and lubrication system uses low-molecular, easily migratable types, it will migrate to the surface first, and after baking, it quickly forms that invisible mist in the cleanroom.
If the lubrication system is changed to a high molecular weight or reactive type, the amount of chip formation will decrease accordingly.
Failure 2: The surface resistance becomes increasingly incorrect the more it is wiped. The root cause is usually that the anti-static treatment is applied as a surface coating or a migratory type, and wiping and cleaning wash away the conductive path together.
This situation tests out beautifully during the development phase, but drops as soon as mass production begins—because no one wipes it five thousand times during development.
Failure Three: After baking the same batch of pieces, the foggy appearance varies in depth. This is not 'unstable material'; it is commonly due to uneven dispersion of additives, with areas of high local concentration precipitating first.
Some pieces in the same batch are good while others are foggy; the direction points to mixed material and masterbatching, not to the substrate.
Failure 4: The ends of the elongated part are warped. The root cause is the fiber orientation stacking absorbing moisture, not insufficient rigidity.
It's very simple to verify this judgment: measure the piece before and after humidity adjustment, and if the amount of warping becomes significantly larger after absorbing moisture, then there's no need to change the material.
There is one sentence I want to say separately: Many people think the cleanliness of a vehicle comes from washing, but actually half of it comes from the formula selection.
Cleaning can only remove the surface residue; it cannot stop the material from continuing to migrate outward.
5. Processing and verification, the order cannot be changed
Most customers who fail on the vehicle didn't skip verification; they just did the steps in the wrong order.
The first thing is drying. If the hygroscopic material has excessive moisture, the feed hopper will be chain-cut, manifesting as brittle pieces, surface silver streaks, and late appearance.
Among the material change complaints we handle, there is a particularly typical type: the same batch of material, the same mold—one mold comes out fine while the next one is fragile, with the formula not changed a bit; it's the drying.
Moisture content is the hardest to maintain during the plum rain season in the south. After unpacking, leaving it in the workshop for a few hours can make the moisture rise again. No matter how well the drying is done, leaving it exposed during turnover is still in vain.
Our approach is to confirm with instruments or dew point data before going online, not based on touch.
The second thing is mold temperature. If the mold temperature is low, the glass fibers will get frozen on the surface, forming floating fibers, and the surface roughness will increase immediately. Later, during cleaning and polishing, the fiber ends will be exposed.
The significance of mold temperature for clean parts is greater than for general parts: for general parts, floating fibers are a cosmetic issue, but on carriers, they are the starting point of debris.
The third thing is the cleaning and drying after molding. The parameters for these two processes must be determined together with the material and cannot simply copy the original settings.
The fourth thing is packaging. The packaging materials themselves may also leach, so clean-grade packaging should be used and validated.
It is recommended to arrange the verification sequence in this way:
1. Material level: moisture content, dry and wet state strength retention
2. Surface level: surface resistance, total deposition amount
3. Process window: Compare parts produced under different mold temperatures and different holding pressures
4. Component Level: Re-measurement of groove spacing and positioning holes after humidity adjustment, roughness after cleaning rounds
5. Client: Complete machine assembly and production line dark field inspection
If the previous item is not passed, just move on; the data measured later has no explanatory value.
Why can't the order be changed? Because surface resistance and deposition are affected by the molded surface condition. If the surface isn't locked, adjusting the cleaning cycles will only yield conclusions applicable to that batch, and the results will fluctuate when production is scaled up.
6. Boundaries: In these situations, don't change them for now
This section may be more valuable than the previous few sections because it helps you cut losses before starting work.
First, production lines whose testing capabilities are not yet in place. Precipitation and surface resistance both require specialized equipment. If currently only weighing and visual inspection are used, even if changed, it still won't prove quality, so the methods should be supplemented first.
Secondly, positions that are soaked in strong oxidative chemical solutions for a long time and have cleaning cycles far exceeding 500 times. For these, one needs to look at fluorine-containing materials or specialized high-purity systems; ordinary modified nylon cannot withstand that many cycles.
Third, parts that need to be partially transparent. For locations like observation windows, it is recommended to use a separate design, made from another material. Do not increase the cost of the entire piece just for a single window.
Fourth, the slot spacing tolerance needs to be pressed to the ±0.01 mm level, and the part is required to be integrally molded in plastic. This level already exceeds the range that injection molding plus moisture absorption can maintain, so it should revert to metal or be made as a structural subcomponent.
Writing these four points first is not to discourage you, but to save time — sample order, batch hold-ups, full case returns, tuition will be much higher than if you didn’t change it at the beginning.
7. Material Replacement Risk List
| link; segment; part | What do you want to move? | Points that are easy to overlook |
|---|
| Mold | The shrinkage rate varies with the filling system, and both ends under long conditions need to be recalculated. | Only replace the material without checking shrinkage, slot spacing drifts |
| Dry | Set the window according to the actual measured moisture content; a dehumidifying dryer is essential. | Hot air drying is basically ineffective for water-absorbing materials |
| Humidity control | Forced moisture adjustment and weight determination, then re-measure the dimensions | Estimate the time based on average wall thickness; the thick walls haven't absorbed fully. |
| Material Temperature and Mold Temperature | The long carbon chain material and glass fiber material have different windows and need to be adjusted together. | Copy the original temperature setting |
| Pressure Holding and Demolding | Switch the demolding system to a low migration model | Using the original release agent, it became the source of precipitation |
| Washing and drying | Re-determination of the number of liquid medicine applications, drying temperature, and time | Using old parameters, water stains and ion residues exceed limits |
| Packaging | Change to clean-grade packaging and perform validation | The packaging material itself is a source of pollution |
| Verification order | Material → Surface → Process → Component level → Client | If the previous item fails, just move on. |
8. Proofing and Trial Molding Schedule
Changing materials and testing molds for the vehicle usually involves three rounds, and the steps are not skipped between rounds.
First round - sample comparison: Use your original mold to make 3-5 samples, only checking moisture content, appearance, and the position of short-shot parts. This round does not pursue performance; first confirm whether the material can be filled in.
Keep two samples, mark the batch number and drying parameters, and keep them at least until the end of the second round.
Second round · Surface and dimensions: fix the material, vary the mold temperature and holding pressure, and make two sets of comparison pieces. Check surface resistance, total precipitation, and the groove distance and positioning holes after humidity adjustment.
This round basically determines the mass production parameters. Samples are sealed and stored by batch, and kept for at least three months after mass production stabilizes.
Round 3 · Cleaning and Client: Run the cleaning cycle at the caliber of 500 times, recheck roughness and strength retention midway, and then send it to the client for assembly and dark-field inspection.
Only after this round passes is it recommended to increase the volume. Keep samples sealed to cover the first batch of mass production, making it easy to track the cause.
The timeline for the three rounds needs to be calculated in advance: usually it takes two weeks from the first round to the second round, and the third round depends on your cleaning line schedule, often taking more than a month.
It's okay to cut rounds, but the round that is cut is usually made up in the form of 'batch return and rewash'.
A one-page sheet for technicians to report upwards can be collected like this:
| Scene | Recommended Route | Key indicators | Verification standard | Conditions that need to be confirmed first |
|---|
| Flower basket, box of slices | Low-precipitation system Main body electrostatic path | Total precipitate amount, surface resistance | Thermal desorption Surface resistivity meter | Cleanliness Level and Baking Temperature |
| Frequently cleaned inner parts | Long carbon chain low precipitation system | Retention rate after washing cycles | Cyclic Soaking ISO 527 | Types and Rounds of Medicinal Liquid |
| High-precision slot pitch component | Mineral-filled system | Tolerance of groove distance after humidity adjustment | Coordinate measuring after humidity adjustment | Measurement Conditions and Reference |
| Load-bearing frame | Glass fiber reinforced system | Mechanical retention rate | ISO 527 | Is electrostatic handled separately? |
9. Several Frequently Asked Questions
Question: After changing materials, is it necessary to go through clean certification again?
Answer: Yes. Certification applies to specific grades; both color masterbatch and release agents must be at the same level. If the material is changed, it must be reported again. There is no shortcut for this.
Question: Can surface resistance be addressed by post-treatment coating?
Answer: Yes, but cleaning and wiping will remove the coating as well. For parts used repeatedly, it's more stable to make the channel inside the main body.
Question: How is batch consistency achieved?
Answer: Vehicle customers generally require sample retention for each batch of key grades, with reports on surface resistance, precipitation, and dimensions for each batch. If any item drifts, the entire batch is put on hold. What we do is a data chain, not a price list.
Question: Why do the first three sentences set the direction?
Answer: The first sentence distinguishes whether what is being changed is the substrate or the entire system, the second sentence asks about the measurement conditions, and the third sentence confirms whether the verification was actually done. If these three sentences are answered fully, the problem is usually narrowed down to one or two items.
The auxiliary system in the formula is matched according to the working conditions per item — conventional auxiliaries are kept in stock, and special models are matched as needed; you report the working conditions and grade, and the materials and auxiliaries are prepared together at once.
About us, four sentences——the matter of choosing materials, the earlier you ask, the less trouble it is.
One is for making modified nylon, one is for mixing materials and additives together, one is to first ask about the working conditions before discussing the price, and one dares to say which parts we don’t do.
The material replacement and mold trial for this type of part can be discussed together.
