今年三月，一家做点火线圈的客户发来一箱外壳。
改性尼龙件，黑色，一个一个剖开看，壳壁内侧有一层很细的裂纹，像瓷器上的冰裂纹，从灌封体边缘往外爬。
他的原话：
>"料是上个月换的，环氧还是同一批环氧。换完之后壳就裂了——到底是料的事，还是胶的事？"
我回了三句：裂在壳上还是裂在灌封体上？冷热循环多少轮开始出现？壳体长期温度多少？
三句问完，答案基本落在第二句里——开裂出现的轮次，是区分"材料问题"和"界面问题"的分水岭。
点火线圈外壳材料这件事，最容易被忽略的一条恰恰是"相容"两个字：它是两个材料之间的事，不是某一个材料的事。
一、工况六维：这里有一条别人不算的温度
温度要看两个值。 线圈本体的长期温度常见在 120–150℃，靠近排气侧瞬态能到 180℃。
传感器罩盖相对温和些，长期多在 100–130℃，但它对尺寸更敏感，因为它管的是信号。
还有一条常被漏掉的温度——固化放热。 环氧固化本身是放热反应，大截面灌封体内部可以比环境高出 30℃ 以上。
也就是说，件在灌封的那一小时里经历的，可能比它一辈子遇到的工况温度还高。
介质是双面的。 外面有溅水、盐雾、机油、清洗剂；里面是灌封胶，以及胶里的胺类固化剂与增塑组分。
这两个面经常被当成一回事，其实它们对材料的要求方向相反。
外面希望材料耐介质、别被吃掉；里面希望材料和胶"黏得住、贴得牢"。
载荷是热循环。 常见考核区间是 -40℃ 到 150℃，一轮几十分钟，寿命期内上千轮。
1000 轮是什么概念？按一年 2 万公里、每 200 公里一次冷启动估，一年大约 100 轮，十年刚好凑够 1000 轮。
寿命看的是绝缘保持。 判据不是"什么时候裂"，是"到寿命终点时，绝缘电阻和耐压还剩多少余量"。
外观与电气。 无裂纹、无析出，同时要过 CTI、绝缘电阻与耐压试验。传感器罩盖还要卡安装尺寸。
一个数字换算：热膨胀系数差多少会出事？
环氧的线膨胀系数大约 50×10⁻⁶/K，玻纤增强尼龙在 40–60×10⁻⁶/K 量级。两者接近时界面应力最小。
如果差出一个量级，几百轮热循环就足以在界面上把裂纹拉出来——这跟强度高低没有关系。
二、材料路线：灌封件和结构件的取舍不一样
| 路线 | 长期耐温 | 吸水率量级 | 灌封相容特点 | 代价 |
|---|
| PA66-GF30 | 130–150℃ | 约 2.5% | 界面结合好，需控小分子 | 吸湿尺寸漂移较大 |
| PA46-GF30 | 150–170℃ | 比 PA66 更高 | 高温下模量保持好 | 吸湿更敏感、工艺窗口窄 |
| PBT-GF30 | 120–140℃ | 约 0.1% | 尺寸稳、收缩小 | 缺口敏感、低温韧性偏弱 |
| PA6T-GF30 | 150℃ 以上 | 约 2%–3% | 耐热余量大 | 成本高、模温要求高 |
| PPS-GF40 | 200℃ 量级 | 约 0.1% 量级 | 尺寸与耐热都好 | 成本高、韧性低 |
看这张表，重点不在"谁强"，在取舍方向。
如果件以"尺寸稳定 + 灌封后不裂"为先，PBT 这一类低吸水体系天然有优势。
如果件以"高温下刚性不能掉"为首要，PA46 或 PA6T 的方向更对。
如果件是结构件、要承受装配与振动，聚酰胺体系的韧性和抗冲击是它的长处。
传感器罩盖和点火线圈外壳，常常不该用同一个料号——一个偏尺寸，一个偏刚性。
三、选型判据表（这一页最该收藏）
| 指标 | 方向性门限 | 验证方法 / 标准 | 常见失效 | 通行解法 | 对应助剂体系 |
|---|
| 长期热氧保留率 | 150℃×1000h 后拉伸保留 ≥70% | ISO 527-2 / GB/T 1040.2-2022 | 表面发黄、脆断 | 稳定化体系按温度档位选 | 抗氧剂（受阻酚 + 亚磷酸酯复配） |
| 热循环后界面 | -40–150℃×1000 轮无可见裂纹 | 件级热循环 + 剖检/超声 | 壳壁冰裂纹、分层 | 匹配线膨胀系数 + 表面处理 | 润滑剂（低迁移型） |
| 界面结合（灌封） | 剪切强度按件定，剖检无弱边界层 | 件级剪切 / 剖检 | 界面剥离、白化 | 灌封前表面活化 + 控脱模剂 | 润滑剂（低迁移型） |
| 吸水率 / 尺寸稳定 | 热循环前吸水饱和尺寸变化 ≤0.3% | ISO 294 / 调湿前后实测 | 装配干涉、间隙漂移 | 低吸水基材或矿物填充 | — |
| 固化收缩与翘曲 | 灌封后平面度按件图纸 | 三坐标 | 壳体翘、灌封偏心 | 浇口与取向设计 | 偶联剂（纤维 / 树脂界面） |
| CTI | 带电件通常要求 ≥600 V | GB/T 4207-2022 / IEC 60112 | 爬电碳化、绝缘失效 | 阻燃与耐漏电体系一起选 | — |
| 绝缘电阻 / 耐压 | 按件规范，寿命末端留余量 | 件级耐压试验 | 击穿、漏电 | 控含水率 + 清洁度 | — |
| 耐介质（外表面） | 机油、盐雾、清洗剂浸泡后无异常 | GB/T 11547-2008 | 表面龟裂、失光 | 结构遮蔽 + 表面防护 | — |
怎么用这张表：先看第二行与第三行——热循环后界面、界面结合。
这两行是这个件最容易翻车的地方，而且它们跟拉伸强度基本没关系。
提醒一句：界面类的项目，一定要拿灌封好的成品件做，不要拿裸壳做。裸壳过得再漂亮，也回答不了界面那一问。
四、四条常见失效，和它们的真实根因
失效一：灌封体边缘爬出细裂纹，壳本身不脆。
大多数情况下，这不是耐温不够，是线膨胀系数差 + 固化收缩在界面上反复做功。
件在热循环里被反复拉压，裂纹从应力最集中的边缘起头，往壁厚方向爬。
这里要否定一个常见做法："加点增韧剂就不会裂"——这个判断是错的。
增韧会降低壳体模量，壳体更软，界面的相对位移反而更大；同时增韧体系与灌封胶的相容性还要另做验证。
正确的方向是先调界面与结构，再谈配方。
失效二：界面发白、剥离，剖开后看到一层"没黏上"的边界。
这往往不是胶的问题，是壳体内壁的脱模剂或润滑剂迁移到表面，形成了一层弱边界层。
助剂侧的一条归因：内壁那一层白，常常是润滑剂与脱模组分往外迁移的结果。换个说法：不是胶没黏住，是内壁上先"上了一层蜡"。
解法是换低迁移体系，并在灌封前加一道表面活化——等离子或火焰处理都行，重点是让内壁"新"起来。
失效三：灌封体内部有气泡、空洞。
先查两件事：灌封工艺的真空脱泡，以及壳体灌封前有没有预烘。
吸湿的聚酰胺件在固化温度下会把水汽放出来，气体只能往灌封体里走，走不掉就成气泡。
这一条是吸湿件独有的坑——PBT 体系在这件事上要老实得多，但它的低温韧性是另一笔账。
失效四：同一批件黄得深浅不一。
这不是"料不稳定"。更可能的原因在混料那一端——抗氧剂或色母没被混匀，颗粒之间就有了色差。
遇到这种件，先去查混料工艺和母粒化那一步，别急着动牌号。
五、加工与验证：顺序要按"界面优先"排
干燥。 聚酰胺必干燥。干燥不到位，料筒里就降解，件出来强度已经打了折。
灌封前预烘。 已经吸湿的壳体，灌封之前要按件壁厚预烘一遍，把水分赶出去再灌。
表面处理。 内壁有脱模剂残留或长期存放的吸附层，灌封前要做活化处理，这一道工序经常被"为了省时间"省掉。
模温。 聚酰胺的结晶度靠模温撑起来。模温低，件脆、表面发暗，耐热也达不到标称值。
验证顺序建议这样排：
1. 调湿后尺寸与平面度（干态数据只做过程记录）
2. 裸件耐热与力学保留（先确认基材站得住）
3. 灌封件热循环（模拟实际界面）
4. 剖检或超声，看界面与内部空洞
5. CTI、绝缘电阻、耐压
6. 整机装车与耐久
顺序不能换。界面这一关不过，后面的电气数据都要重新来。
一个内行细节：热循环试验的冷却段比升温段更值得看。
降温时壳体收缩快、灌封体收缩慢，界面被拉得最狠。只看升温段的方案，往往会漏掉这一处。
六、边界：什么时候这个件不该走改性尼龙
其一，灌封后要求 -40–150℃ 两千轮以上无裂纹，且壁厚在 1.5 毫米以下。
薄壁 + 高轮次，界面的相对位移会被放大，常规聚酰胺体系的余量不足。这类要求要看低膨胀体系或者结构补偿方案。
其二，长期温度超过 180℃。 这个区段里常规改性尼龙的长期性能保持支撑不足。
其三，件要求极低吸水与极小固化后变形。 聚酰胺的吸湿是本性，靠配方压不没。这类件要往低吸水体系走。
其四，年用量小到摊不平模具与长周期验证。 热循环验证本身就要跑上千轮，时间成本不低。
补一节：界面这件事，怎么提前验出来
界面失效的特点是"来得晚、说得清、改不动"。所以要提前验，不能等它爆。
其一，做开壳验。 灌封固化之后，沿灌封体边缘剖一个截面，看界面有没有白线。
其二，做"热循环 + 剖检"。 一轮热循环之后剖一次，比跑完上千轮再剖，更早看出趋势。
其三，要胶厂的三份数据： 线膨胀系数、固化放热曲线、以及它对基材的底涂建议。
其四，常被忘掉的一件——灌封前后量壳体尺寸。 固化收缩会把薄壁往里拉一点，这一点就是装配超差。
| 验证项 | 用什么件 | 看什么 | 出问题时的调整方向 |
|---|
| 界面剖检 | 灌封成品 | 界面有无白线、分层 | 内壁活化、控脱模剂 |
| 热循环剖面 | 灌封成品 | 裂纹起头位置与出现轮次 | 匹配线膨胀系数 |
| 固化放热 | 灌封成品（埋热电偶） | 灌封体内部峰值温度 | 分批灌、控单次胶量 |
| 壳体尺寸 | 灌封前后对比 | 壁厚与平面度变化 | 调整固化曲线 |
| 绝缘 | 灌封成品 | 耐压与绝缘电阻 | 控含水率与清洁度 |
再补一条时间线，这类失效的典型路径就是这样：
`
换料 ├── 灌封正常，首件检验合格，外观无异常
│
第 3 个月
├── 售后偶发不识别（当时判为端子问题）
├── 个别件壳内有轻微发白（当时判为清洗残留）
└── 第 11 个月批量壳壁细裂 → 剖面见界面白线 → 改表面活化 + 换低迁移体系
`
回头看，最早的信号在第 3 个月就出现了。 只是它长得像另一个问题。
所以界面验证的价值不在于"证明没问题"，在于把信号的时点提前。
换料风险清单（从原方案换过来，要动的几项）
| 环节 | 要动什么 | 容易漏的点 |
|---|
| 模具 | 收缩率随玻纤含量变，壳体与盖要同步修 | 灌封位壁厚与圆角 |
| 干燥 | 按实测含水率定窗口，用除湿干燥机 | 回用料掺入带进的水分 |
| 灌封前预烘 | 按件壁厚定时长，赶水后再灌 | 直接拿吸湿件灌封 |
| 表面处理 | 内壁活化，去除脱模剂残留 | 为了省时间省掉这一道 |
| 料温 / 模温 | 模温按结晶度与表面质量联合调 | 只抄牌号推荐值，不看件 |
| 灌封工艺 | 真空脱泡参数要重定 | 沿用原方案的固化曲线 |
| 界面验证 | 必须用灌封好的成品件做热循环 | 用裸壳数据代替界面数据 |
| 验证顺序 | 尺寸 → 耐热 → 界面 → 电气 → 整机 | 前一项未过就往下走 |
一页纸汇报表（给要向上汇报的人）
`
项目：点火线圈外壳 / 传感器罩盖 · 材料路线评估
结论方向：外壳与罩盖建议分开选材，界面验证是核心关口
一、必须守住的三条
1. 界面验证用灌封好的成品件做，不接受裸壳数据
2. 灌封前预烘 + 内壁活化，两道工序进作业标准
3. 热循环看冷却段的界面表现，不只记录峰值
二、前置条件（任一不满足则建议暂缓）
· 长期温度在 180℃ 以内
· 壁厚与热循环轮次在聚酰胺体系可覆盖范围内
· 有灌封设备与热循环验证资源
· 壳体与灌封胶的相容性数据可从胶厂拿到
三、下一步动作
1. 取灌封胶的线膨胀系数与固化放热曲线
2. 做一轮 -40–150℃ 热循环，剖检界面
3. 测灌封前后壳体尺寸，评估固化收缩影响
风险提示：本路线的主要不确定性在界面长期表现，不在初始强度与耐热。
`
读者常问的两句
问：和进口料比，国产路线差在哪？
按公开资料口径，进口牌号在这类件上的优势主要是长期热老化数据链完整、批次稳定性记录齐全，以及与主流灌封胶体系的配套验证经验。
国产路线的差距更多在"配套数据齐不齐"这件事上。哪些件上已经成熟、哪些件仍不建议，要看件的验证结果，不能一概而论。
问：传感器罩盖能不能去掉玻纤，换更好的密封性？
方向要对。去玻纤，件会软，罩盖的卡装结构和平面度都容易失守；密封靠的是结构与尺寸，不是靠材料变软。
如果是为了尺寸稳定，正确做法是用矿物填充改善各向异性，而不是把玻纤全部拿掉。
问：灌封胶能不能换？壳体能跟着不换吗？
方向是"要一起验"。换胶等于换了一个界面体系：固化曲线、线膨胀系数、固化放热都可能变。
壳体的表面状态没变，但界面能不能贴牢会变。换胶之后至少要重做一轮热循环加剖检，不能只比一份胶的物性表。
结语
回到开篇那三句问话。为什么这三句能把方向定下来？
因为它问的是三件事：问裂在哪（定是材料还是界面）、问多少轮（定是短期还是长期）、问温度（定材料体系）。
这三句问完，才轮到牌号出场。
如果你手上正有点火线圈外壳或传感器罩盖要定料，把三样东西发过来就能给方向：长期工作温度、灌封胶的体系与固化曲线、热循环轮次要求。
这些年被问得最多的一句，大概是「副牌料到底能不能用」。
答案从来不是"能"或"不能"，是偏在哪一项、被用在哪个件上——灌封件这类涉及界面与安规的件，我们一般不建议拿它试。
我们做的事很具体：把 PA6、PA66、PA46、PA11、PA12、PA6T、PA9T 和尼龙合金这些树脂，改成某个件真正能用的样子；顺带做改性 PPO、PPS 和热塑性弹性体。
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。
也经营各大化工巨头的尼龙树脂、副牌料和大包料，另长期收尼龙原料、水口回料与各类尼龙废料，有正规处置渠道。
灌封件这类件的选料与试模，可以一起聊。
This March, a customer who makes ignition coils sent a box of housings.
Modified nylon parts, black, examined one by one, with a very fine layer of cracks on the inside of the shell wall, resembling crazing on porcelain, extending outward from the edge of the encapsulation body.
His exact words:
The material was replaced last month, and the epoxy is still from the same batch. After the replacement, the shell cracked—so is it the material's problem or the glue's problem?
I replied with three questions: Is the crack on the shell or on the potting? After how many thermal cycles does it start to appear? What is the long-term temperature of the shell?
After asking three questions, the answer basically lies in the second one—the round in which cracking occurs is the dividing line between a 'material problem' and an 'interface problem'.
Regarding the material of the ignition coil housing, the easiest point to be overlooked is precisely the word 'compatibility': it concerns the relationship between two materials, not any single material.
1. Six-dimensional working condition: Here is a temperature that others don’t count
Temperature should look at two values. The long-term temperature of the coil body is commonly 120–150°C, and near the exhaust side it can transiently reach 180°C.
The sensor cover is relatively mild, usually staying at 100–130°C for long periods, but it is more sensitive to size because it handles signals.
There is another commonly overlooked temperature—the heat released during curing. Epoxy curing itself is an exothermic reaction, and the interior of a large-section encapsulation can be more than 30°C higher than the ambient temperature.
In other words, the conditions experienced by the component during the one hour of potting could be higher than the operating temperatures it encounters throughout its entire life.
The medium is double-sided. The outside is exposed to splashes, salt spray, engine oil, and cleaning agents; the inside contains potting compound, as well as amine curing agents and plasticizing components in the adhesive.
These two aspects are often treated as the same thing, but in fact, they have opposite requirements for the material.
On the outside, we hope the material is resistant to the medium and won’t be eaten away; on the inside, we hope the material can stick well with the adhesive and bond firmly.
The load is thermal cycling. The common evaluation range is from -40℃ to 150℃, with each cycle lasting dozens of minutes, and thousands of cycles over the service life.
What does 1000 cycles mean? Assuming 20,000 kilometers per year and a cold start every 200 kilometers, that's about 100 cycles per year, so ten years would just make up 1000 cycles.
Lifespan is determined by the insulation retention. The criterion is not 'when it cracks,' but 'how much margin of insulation resistance and withstand voltage remains at the end of its life.'
Appearance and electrical. No cracks, no precipitation, and it must pass CTI, insulation resistance, and withstand voltage tests. The sensor cover also needs to fit the installation dimensions.
A numeric conversion: How much difference in the thermal expansion coefficient will cause problems?
The linear expansion coefficient of epoxy is about 50×10⁻⁶/K, while glass fiber reinforced nylon is on the order of 40–60×10⁻⁶/K. When the two are close, the interfacial stress is minimal.
If the difference reaches an order of magnitude, a few hundred thermal cycles are enough to pull cracks out at the interface — this has nothing to do with the strength.
2. Material route: The selection of encapsulated parts and structural parts is different
| Route | Long-term heat resistance | Water absorption magnitude | Potting compatibility characteristics | Cost |
|---|
| PA66-GF30 | 130–150℃ | About 2.5% | The interface is well integrated, small molecules need to be controlled | Moisture-induced dimensional drift is relatively large |
| PA46-GF30 | 150–170°C | Higher than PA66 | Maintains good modulus at high temperatures | More sensitive to moisture, narrow processing window |
| PBT-GF30 | 120–140°C | About 0.1% | Stable size, low shrinkage | Notch sensitive, low-temperature toughness is relatively weak |
| PA6T-GF30 | Above 150℃ | About 2%–3% | Large heat resistance margin | High cost, high mold temperature requirements |
| PPS-GF40 | 200℃ scale | About 0.1% level | The size and heat resistance are both good | High cost, low toughness |
Looking at this table, the focus is not on 'who is stronger,' but on the direction of trade-offs.
If the priority is 'dimensional stability, no cracking after potting,' PBT and similar low water-absorption systems naturally have an advantage.
If the priority of the part is 'maintaining rigidity at high temperatures,' PA46 or PA6T would be more suitable.
If the part is a structural component and needs to withstand assembly and vibration, the toughness and impact resistance of the polyamide system are its strengths.
The sensor cover and the ignition coil housing often should not use the same part number—one tends to be off in size, the other in rigidity.
3. Selection Criteria Table (This page is the most worth keeping)
| Indicator | Directional Threshold | Verification Method / Standard | Common Failures | Common solution | Corresponding auxiliary agent system |
|---|
| Long-term thermal oxygen retention rate | After 150℃ × 1000h, tensile retention ≥ 70% | ISO 527-2 / GB/T 1040.2-2022 | Yellowed surface, brittle and broken | Select the stabilization system according to the temperature setting | Antioxidant (hindered phenol and phosphite blend) |
| Interface after thermal cycling | -40–150℃ × 1000 cycles without visible cracks | Component-level thermal cycling Sectioning/Ultrasound | Shell wall ice cracks and delamination | Matching linear expansion coefficient Surface treatment | Lubricant (low migration type) |
| Interface bonding (potting) | Shear strength is specified per piece, and no weak boundary layer is found in the inspection. | Component-level Shearing / Sectioning | Interface stripping, whitening | Surface activation before potting Control release agent | Lubricant (low migration type) |
| Water Absorption / Dimensional Stability | Dimensional change after water absorption before thermal cycling ≤0.3% | ISO 294 / Measured Before and After Humidity Adjustment | Assembly interference, clearance drift | Low water-absorbent substrate or mineral filler | — |
| Curing shrinkage and warping | After potting, the flatness is according to the individual drawing | CMM | Case warping, eccentric potting | Gate and Orientation Design | Coupling Agent (Fiber / Resin Interface) |
| CTI | Live parts usually require ≥600 V | GB/T 4207-2022 / IEC 60112 | Tracking carbonization, insulation failure | Flame-retardant and leak-proof systems are selected together | — |
| Insulation Resistance / Withstand Voltage | According to item specifications, leave a margin at the end of the lifespan | Component-level pressure resistance test | Breakdown, leakage | Control moisture content Cleanliness | — |
| Medium-resistant (outer surface) | No abnormalities after soaking in engine oil, salt spray, and cleaning agents | GB/T 11547-2008 | Surface cracking, dullness | Structural shielding Surface protection | — |
How to use this table: First look at the second and third rows — interface after thermal cycling, interface bonding.
These two lines are the places where this part is most likely to fail, and they have basically nothing to do with tensile strength.
A reminder: For interface-type projects, you must use fully assembled finished products, not just the bare shells. No matter how beautiful the bare shell is, it won't answer any questions about the interface.
4. Four common failures and their real root causes
Failure 1: Fine cracks appear along the edges of the potting body, while the shell itself is not brittle.
In most cases, this is not due to insufficient temperature resistance, but a difference in the coefficient of thermal expansion, with curing shrinkage repeatedly doing work at the interface.
The part is repeatedly stretched and compressed in the thermal cycle, and cracks start from the edge where the stress is most concentrated and creep in the direction of the wall thickness.
Here, a common practice needs to be negated: 'Adding some toughening agent will prevent cracking'—this judgment is wrong.
Toughening will reduce the modulus of the casing, making the casing softer, and the relative displacement at the interface will actually be larger; at the same time, the compatibility between the toughening system and the potting adhesive also needs to be further verified.
The correct approach is to first adjust the interface and structure, and then discuss the formulation.
Failure 2: The interface appears white and peels off. When cut open, a boundary that 'did not bond' is seen.
This is often not a problem with the adhesive; it is the release agent or lubricant on the inner wall of the casing that migrates to the surface, forming a weak boundary layer.
One attribution from the additive side: the white layer on the inner wall is often the result of lubricants and release agents migrating outward. In other words: it's not that the glue didn't stick, but that a 'layer of wax' was applied on the inner wall first.
The solution is to switch to a low-migration system and add a surface activation step before filling—either plasma or flame treatment works; the key is to make the inner wall 'new' again.
Failure three: There are bubbles or voids inside the potting compound.
Check two things first: the vacuum degassing of the potting process, and whether the housing has been pre-baked before potting.
Moist polyamide parts will release water vapor at curing temperature, and the gas can only move into the encapsulant; if it cannot escape, it turns into bubbles.
This is a pit unique to hygroscopic materials — the PBT system is much more honest in this regard, but its low-temperature toughness is another matter.
Failure 4: The same batch of items has inconsistent yellowing depth.
This is not an 'unstable material.' The more likely reason is on the mixing end—if the antioxidant or color masterbatch is not mixed evenly, there will be color differences between the pellets.
When encountering this kind of piece, first check the blending process and the masterbatch step, don't rush to change the grade.
5. Processing and Verification: The order should follow 'interface priority'
Drying. Polyamide must be dried. If it is not dried properly, it will degrade in the barrel, and the strength of the molded parts will already be compromised.
Pre-baking before potting. For cases that have already absorbed moisture, they need to be pre-baked according to the wall thickness before potting, to drive out the moisture before potting.
Surface treatment. The inner wall may have mold release residue or an adsorbed layer from long-term storage, so activation treatment must be done before potting. This step is often 'skipped to save time'.
Mold temperature. The crystallinity of polyamide is supported by the mold temperature. If the mold temperature is low, the part becomes brittle, the surface darkens, and the heat resistance does not reach the nominal value.
It is recommended to arrange the verification sequence in this way:
1. Dimensions and flatness after humidity adjustment (dry state data is recorded only for process tracking)
2. Heat resistance and mechanical retention of bare parts (first confirm that the substrate can withstand it)
3. Potting component thermal cycling (simulate actual interface)
4. Dissection or ultrasound, observe the interface and internal cavities
5. CTI, Insulation Resistance, Dielectric Withstand Voltage
6. Complete Vehicle Installation and Durability
The order cannot be changed. If you fail this interface stage, all the subsequent electrical data will have to be redone.
A professional detail: the cooling phase of a thermal cycling test is more worth watching than the heating phase.
When cooling, the casing shrinks quickly while the potting compound shrinks slowly, putting the most strain on the interface. If you only look at the heating phase of the plan, this part is often overlooked.
6. Boundaries: When this part should not use modified nylon
First, after potting, it is required to withstand more than 2,000 cycles from -40 to 150°C without cracks, and the wall thickness should be less than 1.5 millimeters.
Thin-walled: At high cycles, the relative displacement of the interface will be amplified, and the margin of conventional polyamide systems is insufficient. Such requirements need to be addressed by low-expansion systems or structural compensation solutions.
Secondly, the long-term temperature exceeds 180℃. In this range, the long-term performance retention of conventional modified nylon is insufficient.
Third, the parts require very low water absorption and minimal deformation after curing. Moisture absorption is inherent to polyamides and cannot be completely suppressed by formulation. Such parts need to move towards a low water absorption system.
Fourth, the annual usage is too small to justify spreading the costs of molds and long-term validation. Thermal cycle validation itself requires running thousands of cycles, and the time cost is not low.
Adding a section: The matter of the interface, how to detect it in advance
The characteristics of an interface failure are 'it comes late, it is explainable, but it cannot be fixed.' Therefore, it must be tested in advance and cannot wait until it breaks.
First, perform a shell-opening inspection. After the potting has cured, cut a cross-section along the edge of the potting body to see if there is a white line at the interface.
Second, perform a 'thermal cycle dissection.' After one round of thermal cycling, dissect once; this allows trends to be observed earlier than running thousands of cycles before dissecting.
Third, three pieces of data are required from the glue factory: the linear expansion coefficient, the curing exotherm curve, and its primer recommendations for the substrate.
Fourthly, one often forgotten thing—measuring the shell dimensions before and after potting. Curing shrinkage will pull the thin walls inward slightly, and this slight difference is the assembly tolerance.
| Verification item | Using what part? | What are you looking at? | Direction of adjustments when problems arise |
|---|
| Interface sectioning | Encapsulated finished product | Whether there are white lines or layering on the interface | Inner wall activation, release agent control |
| Thermal cycling profile | Encapsulated finished product | Crack initiation location and occurrence cycle | Matching coefficient of linear expansion |
| Exothermic curing | Potted finished product (embedded thermocouple) | Peak temperature inside the encapsulant | Gradually pour and control the amount of glue per application |
| Housing size | Comparison before and after potting | Wall thickness and flatness variation | Adjust the curing curve |
| Insulation | Encapsulated finished product | Withstand voltage and insulation resistance | Control of moisture content and cleanliness |
Adding another timeline, the typical path of such failures is like this:
`
Material change ├── Encapsulation is normal, first-piece inspection passed, appearance has no abnormalities
│
Month 3
├── Occasional unrecognition after-sales (at that time judged as a terminal issue)
├── Some individual cases have slight whitening inside the casing (at that time, it was judged to be washing residue)
└── Batch shell wall fine cracks in the 11th month → Cross-section shows interface white line → Modify surface activation Switch to low-migration system
`
Looking back, the earliest signal appeared in the third month. It just looked like another problem.
So the value of interface verification is not in 'proving that there is no problem,' but in advancing the timing of the signal.
Material Change Risk List (Transferred from the original plan, items that need to be changed)
| link; segment; part | What do you want to move? | Points that are easy to overlook |
|---|
| Mold | The shrinkage rate changes with the glass fiber content, and the housing and cover need to be adjusted simultaneously. | Wall thickness and fillets of the potting position |
| Dry | Set the window according to the measured moisture content, using a dehumidifying dryer | Recycled materials mixed with the water content brought in |
| Pre-baking before potting | Set the time according to the wall thickness of each piece, and pour after removing the water. | Directly use the desiccant for potting |
| Surface treatment | Inner wall activation, removal of release agent residue | To save time by skipping this step |
| Material Temperature / Mold Temperature | Mold temperature is adjusted jointly according to crystallinity and surface quality | Only copy the recommended brand numbers, do not look at the parts. |
| Encapsulation process | The vacuum degassing parameters need to be reset | Retain the original plan's fixed curve |
| Interface verification | Thermal cycling must be done using fully encapsulated finished parts. | Use raw shell data instead of interface data |
| Verification order | Size → Heat Resistance → Interface → Electrical → Complete Machine | If the previous item fails, just move on. |
One-page report form (for people who need to report upwards)
`
Project: Ignition Coil Housing / Sensor Cover · Material Route Assessment
Conclusion direction: It is recommended to select materials for the shell and cover separately, and interface verification is the key checkpoint.
1. Three Rules That Must Be Followed
1. Use fully assembled and encapsulated finished products for interface verification; bare shell data is not accepted.
2. Pre-baking before potting Inner wall activation, two-step process operation standard
3. In thermal cycling, observe the interface performance of the cooling section, not just record the peak values
2. Precondition (It is recommended to postpone if any are not met)
· Long-term temperature within 180℃
· Wall thickness and thermal cycling rounds are within the coverable range in the polyamide system
· Equipped with potting equipment and thermal cycle verification resources
· Compatibility data between the casing and the potting glue can be obtained from the adhesive manufacturer
3. Next Steps
1. Obtain the coefficient of linear expansion of the potting compound and the curing exotherm curve
2. Perform a cycle of -40–150℃ thermal cycling and cross-section the interface
3. Measure the shell dimensions before and after potting to assess the impact of curing shrinkage
Risk Warning: The main uncertainty of this route lies in the long-term interface performance, not in the initial strength and heat resistance.
`
Two questions readers often ask
Question: Compared with imported materials, where does the domestic route fall short?
According to publicly available information, the advantage of imported grades in this type of part mainly lies in the completeness of long-term thermal aging data chains, the full records of batch stability, and the experience in compatibility verification with mainstream potting adhesive systems.
The gap in domestic routes is more about whether supporting data is complete. Which parts are already mature and which are still not recommended depends on the test results of the parts, and cannot be generalized.
Question: Can the sensor cover remove the fiberglass and switch to better sealing?
The direction must be correct. If you go with fiberglass, the part will become soft, and the snap-fit structure and flatness of the cover are easily compromised; sealing relies on structure and dimensions, not on the material softening.
If it is for dimensional stability, the correct approach is to use mineral fillers to improve anisotropy, rather than removing all the fiberglass.
Question: Can the potting adhesive be replaced? Can the housing remain unchanged?
The direction is to 'verify together.' Changing the adhesive is equivalent to changing an interface system: the curing curve, coefficient of linear expansion, and curing heat release may all change.
The surface condition of the casing hasn't changed, but whether the interface can bond firmly will change. After changing the adhesive, at least one round of thermal cycling plus cross-sectional inspection must be redone; you can't just compare the property table of one adhesive.
Conclusion
Returning to the three questions at the beginning. Why can these three questions set the direction?
Because it asks three things: where the crack is (determining whether it is the material or the interface), how many cycles (determining whether it is short-term or long-term), and the temperature (determining the material system).
After these three questions are asked, it's finally the turn for the brand to appear.
If you currently have a few ignition coil housings or sensor covers to specify materials for, just send over these three items, and we can provide guidance: long-term operating temperature, the system and curing curve of the potting adhesive, and the required number of thermal cycles.
The question I've been asked most over the years is probably 'Can secondary leather actually be used?'
The answer has never been 'can' or 'cannot'; it depends on which part it is偏 and which component it is used on—components like encapsulated parts that involve interfaces and safety standards, we generally do not recommend using it for testing.
What we do is very specific: we take resins like PA6, PA66, PA46, PA11, PA12, PA6T, PA9T, and nylon alloys, and turn them into a form that can actually be used for a certain part; we also do modified PPO, PPS, and thermoplastic elastomers along the way.
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.
Also operates the nylon resin, secondary-grade materials, and bulk materials of major chemical giants, and has long-term purchasing of nylon raw materials, sprue back materials, and various types of nylon waste, with proper disposal channels.
We can discuss the material selection and mold testing for potting parts together.
