去年秋天，一家做热管理模块的客户寄来两批节温器壳体。
改性尼龙件，玻纤增强，同一牌号、同一副模具、同一份工艺卡，出自同一家厂的两个班。
一批装车跑了十八个月，法兰面开始渗冷却液；另一批快三年了，拆下来还是干的。
客户的原话很直接：
>"料是同一个料，我们也是照着他那样打的。为什么他那批裂，我们不裂？"
我回了三句：长期水温多少？乙二醇浓度多少？法兰螺栓的预紧力多少？
三句问完，路径就出来了。节温器壳体材料这道题，答案常常不在牌号上，在"温度、介质、预紧力"这三个数字的组合里。
一、工况六维：这里有两个变量别人不常算
温度要看长期值。 节温器壳体长期泡在 105–130℃ 的冷却液里，热管理模块支架靠排气的那一侧还能更高。
节温器本身是频繁动作件，开闭周期常见在两三分钟量级——它一直在冷热交变里过日子。
130℃ 的水是什么概念？家用高压锅工作时大约 120℃。而这个件要在里面连续泡上万小时。
介质不是"乙二醇"，是"冷却液"。 典型配比是 50% 乙二醇加 50% 水，再加缓蚀剂、消泡剂。
这里有一条关键的化学变化：乙二醇在高温和有氧条件下会氧化，生成乙醇酸、乙二酸一类的小分子酸。
结果就是冷却液的 pH 会从新液时的 8 上下，慢慢往下降。pH 掉到 6 附近的时候，对聚酰胺就不客气了。
压力是交变的。 系统压力常见在 0.1–0.2 MPa 量级，随水温上下波动，一天里反复多次。
载荷里有一条是"静"的。 法兰螺栓的预紧力是长期压在壳体上的，这一条带来的是蠕变，不是强度问题。
寿命按整车算。 15 年 / 24 万公里，期间热循环次数是数万量级。
外观与合规。 密封面平面度、内流道通畅度、耐冷却液后的性能保留，这三项要在定点前定清口径。
两个数字换算：
一是热循环次数。按一天两次冷启动、一年 365 天算，一年七百次上下，十五年就是一万次量级。
二是蠕变。130℃ 下持续受压的塑料件，1000 小时后的形变量，往往比常温下同样载荷的形变量高一个量级。
这第二个数字，才是法兰渗漏的真正来源。
二、材料路线：耐水解和耐热不是一回事
| 路线 | 长期耐温 | 耐水解体系 | 吸水率量级 | 代价 |
|---|
| PA6-GF30 | 100–120℃ | 弱 | 约 2.8–3.0% | 长期 130℃ 水解风险高 |
| PA66-GF30 | 130–150℃ | 成熟（热稳定体系可配） | 约 2.5% | 吸湿尺寸漂移较大 |
| PA66-GF35 | 130–150℃ | 成熟 | 约 2.2% | 熔接线强度随玻纤上升下降 |
| PA6T-GF30 | 150℃ 以上 | 中上 | 约 2%–3% | 成本高、模温要求高 |
| PPS-GF40 | 200℃ 量级 | 好 | 约 0.1% 量级 | 成本高、韧性低 |
看这张表，重点在"耐热"和"耐水解"是两件事。
一个材料能扛住 150℃ 的短时峰值，不代表它能在 130℃ 的水和乙二醇里泡三年——后面这件事考的是酰胺键的水解稳定性。
PA66 体系在这类件上是主流，前提是配了合适的热稳定与抗水解体系；配与不配，长期表现在同一条线上会分开。
PA6T 的耐温余量更大，代价是工艺窗口与成本；PPS 的尺寸和耐介质都好，代价是韧性与成本。
热管理支架和节温器壳体也常常不该用同一个料号：前者偏结构刚性，后者偏密封与耐水解。
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 |
|---|
| 耐冷却液后力学保留 | 130℃×1000h 后拉伸保留 ≥70% | ISO 527-2 / GB/T 1040.2-2022 | 脆化、表面龟裂 | 抗水解体系按温度档位配 | 抗氧剂（含热稳定复配体系） |
| 分子量保持 | 浸泡后相对黏数下降在可控区间 | GB/T 1632.1 / ISO 307 | 长期性能断崖式下滑 | 控含水率 + 抗水解体系 | 抗氧剂（含热稳定复配体系） |
| 130℃ 蠕变 | 1000h 形变量按件定，重点看法兰 | 恒温恒载蠕变试验 | 法兰渗漏、预紧力松弛 | 玻纤增强 + 结构加强筋 | — |
| 密封面平面度保持 | 热循环后仍在图纸公差内 | 三坐标 / 件级热循环 | 渗漏、密封圈压不实 | 浇口与取向设计 | — |
| 内流道析出物 | 无可见析出，流道截面不缩 | 浸泡后剖检 + 称重 | 阀口卡滞、流量漂移 | 控小分子助剂用量 | 润滑剂（低析出型） |
| 熔接线强度 | ≥ 本体强度的 60% | 件级剖检 + 拉伸 | 熔接线开裂 | 控玻纤含量 + 浇口设计 | 偶联剂（纤维 / 树脂界面） |
| 爆破压力 | 按件规范，留余量 | 液压爆破试验 | 壳体爆裂 | 结构与壁厚配合 | — |
| 吸水后尺寸变化 | 0.3% 量级以内 | ISO 294 / 调湿前后实测 | 装配干涉、密封面偏移 | 强制调湿或低吸水基材 | — |
怎么用这张表：先看头两行——耐冷却液后的力学保留与分子量保持。
这两行代表的是"材料有没有被水吃掉"。它们过不去，后面的蠕变和爆破数据都只是暂时好看。
提醒一句：耐冷却液数据一定要写明冷却液的品类、浓度、pH 与温度。只写"耐乙二醇"四个字的数据，参考价值有限。
四、五条常见失效，和它们的真实根因
失效一：法兰面慢慢渗液，壳体本身不裂。
根因多半是蠕变，不是强度。螺栓预紧力长期压在 130℃ 的塑料上，材料会一点点流走。
预紧力一松，密封圈压不实，渗漏就来了。这一条在常温试验里基本测不出来。
失效二：表面龟裂、粉化，一掰就断。
这是水解加长期热氧。为什么水和酸会这么厉害？
聚酰胺的主链是酰胺键，水和酸都能把它切开——分子链一段段变短，宏观上就是强度垮掉。
而冷却液用久了会变酸，这件事让它从"中性水"变成了"酸性水"，攻击性强了一大截。
助剂侧的一条归因：这类件靠热稳定与抗水解体系吃饭，但有些复配体系在含胺、含硫的冷却液环境里会被络合而失活。同一牌号在不同冷却液里寿命差一截，常常是这件事。
失效三：节温器阀口卡滞，流量漂移。
这不是机械问题，是析出物。小分子助剂被热水抽出来，在阀口和阀座附近慢慢堆积。
排查时先看流道剖检，再看配方里的小分子用量——这一条比换基材更快见效。
失效四：同一批件，一批能用一批漏。
先查三件事：干燥、装配预紧力、冷却液品牌与批次。这三件事里任何一件不一致，都足以拉开寿命差距。
失效五：靠近排气侧的那一角先裂。
这是温度分布不均的典型信号。先在整机上补测温点，再谈材料。
这里有一条要直说： 很多人以为"玻纤加得越多越抗蠕变"。
方向对了一半。玻纤确实能提高初始刚性，但它挡不住基体在吸湿和高温下的模量下降。
而且玻纤含量上去之后，纤维与基体的界面在长期水环境下会成为裂纹的起点，熔接线强度也跟着掉。
所以我们更愿意问一句：这件是"刚性不够"还是"密封面在流"？两件事的解法不一样。
五、加工与验证：这根链条上干燥是最前面的一环
干燥。 这一条我们自己在车间里吃过亏：同一批料、同一副模具、参数一字未动，这一模好好的，下一模一掰就碎。
查了三天模具和参数，什么都没查出来。最后翻到干燥机——那几天用的是热风干燥机。
聚酰胺必须用除湿干燥机，普通热风机对它是无效的。含水率超标的料在料筒里就水解降解，件的强度已经打了折。
再泡到 130℃ 的水里，寿命更短。这类问题在原料袋上测不出来，只在件上显形，而且显形得很晚。
模温。 结晶度靠模温撑起来。模温低，件脆、表面发暗，耐水解表现也跟着降。
浇口与取向。 密封面附近希望纤维取向一致，避免局部收缩差把平面度拉跑。
验证顺序建议这样排：
1. 调湿后尺寸与密封面平面度（干态数据只做过程记录）
2. 130℃ 冷却液浸泡后的力学保留与分子量变化
3. 130℃ 恒温恒载蠕变（看法兰形变量）
4. 内压循环与爆破
5. 件级热循环后的密封试验
6. 整机装车与耐久
顺序不能换。前一项不过，后面测出来的数只是暂时好看。
一个内行细节：蠕变试验的时长不能压缩。
前三个月的形变速率不能外推后三年——塑料的蠕变曲线不是一条直线，它会在某个时间点突然变陡。
六、边界：什么时候该退回金属或换体系
其一，长期水温超过 140℃。 这个区段里常规聚酰胺体系的长期性能保持支撑不足，要往 PPS 一类体系走。
其二，冷却液长期处在酸性区间且无法更换。 如果使用环境的冷却液无法按周期维护，水解这条路的风险会被放大。
其三，要求长期零渗漏且不可维护。 这类件的失效后果是漏水停机，风险权重很高，要看结构补偿方案或者回到金属。
其四，年用量小到摊不平模具与长周期验证成本。
把这四条摆出来不是劝退，是让项目在立项阶段就知道验证要跑多久。
补一节：蠕变验证怎么做，才敢在报告上签字
这一节写给要签字的人。蠕变试验做错了，比不做更危险，因为它会给你一个偏乐观的数。
温度按长期值取。 拿 23℃ 的蠕变曲线去推 130℃ 的形变，结论一定是错的。
载荷按实际预紧力折算，不要只按一半。 保守的载荷得到的是保守的结论，反之亦然。
测点选在法兰，不要选在壳体中间。 你要回答的是"密封面会不会松开"，不是"这块料好不好"。
三个时间点都要记录： 1000 小时、3000 小时、5000 小时。前段平缓，不代表后段也平缓。
同时测预紧力衰减。 螺栓扭矩的衰减曲线，比形变量更接近"渗不漏"这个问题的答案。
| 试验项 | 条件 | 看什么 | 判定方向 |
|---|
| 恒温恒载蠕变 | 长期水温、实际预紧力 | 法兰处形变量 | 形变量决定密封余量 |
| 预紧力衰减 | 同上 | 螺栓扭矩随时间变化 | 决定装配扭矩与复查周期 |
| 浸泡后力学 | 130℃×1000h 冷却液 | 拉伸保留率 | 决定材料是否被水吃掉 |
| 分子量变化 | 同上 | 相对黏数 | 断链程度 |
| 件级热循环 | -40℃ 到 130℃ | 密封面平面度 | 决定密封结构 |
再补一条时间线——同一袋料、两套工艺，差距是怎么拉开的：
`
同一牌号 ├── 两个班、两套干燥与预紧参数，首件都合格
│
第 6 个月
├── 少量件法兰面有白色析出（判为冷却液脏）
├── 冷启动后偶发液位报警（判为传感器）
└── 第 18 个月集中报渗漏 → 追溯 → 两批预紧力差一成半 → 重定扭矩与干燥窗口
`
结论不是"料不行"。 结论是同一袋料在两套工艺下，寿命差了一年半。
这也是为什么这类件的评审，我建议把"工艺与扭矩"写进技术协议：换干燥设备、换螺栓供应商，都要重新确认一次。
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，密封面按件补偿 | 密封槽位置与壁厚突变 |
| 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 |
| 料温 / 模温 | 模温按结晶度与平面度联合调 | 只抄牌号推荐值，不看件 |
| 装配预紧力 | 按材料蠕变特性重定扭矩 | 沿用原金属件的扭矩值 |
| 调湿 | 强制调湿 + 复测尺寸 | 按平均壁厚估时间，厚壁处没吸透 |
| 冷却液验证 | 写明品类、浓度、pH 与温度 | 只做纯乙二醇一组 |
| 结构 | 密封槽与加强筋要一起改 | 只改材料、不动结构 |
| 验证顺序 | 尺寸 → 介质 → 蠕变 → 内压 → 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：节温器壳体 / 热管理模块支架 · 材料路线评估
结论方向：改性尼龙可作候选路线，前提是耐水解体系与结构补偿同时到位
一、必须守住的三条
1. 130℃ 冷却液浸泡数据 + 分子量变化数据，两组都要
2. 装配预紧力按蠕变特性重定，不能沿用金属件扭矩
3. 蠕变验证时长不压缩，不满时长不出结论
二、前置条件（任一不满足则建议暂缓）
· 长期水温在 140℃ 以内
· 冷却液有定期维护与更换机制
· 有恒温恒载蠕变试验条件
· 密封面可通过结构补偿控制
三、下一步动作
1. 实测实际使用环境的冷却液 pH 与温度
2. 做 130℃×1000h 浸泡，看力学保留与分子量
3. 做 130℃ 恒载蠕变，量法兰形变量
风险提示：本路线的主要不确定性在长期水解与蠕变，不在初始强度与耐温。
`
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水解数据的完整性、批次稳定性记录，以及配套冷却液体系的验证经验。
国产路线的差别更多在"长期数据有没有做全"这件事上，不一定在材料本身。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：热管理支架能不能直接沿用节温器壳体的料号？
看它怕什么。支架主要怕刚性和蠕变，壳体主要怕水解和密封。两者的优先级不同。
如果件数不多、管理上想并料号，至少要用支架的工况把蠕变重新验一遍——省下的是管理成本，不能省验证。
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：问水温（定材料体系）、问浓度（定介质严苛程度）、问预紧力（定蠕变风险）。
这三句问完，才轮到牌号出场。
如果你手上正有节温器壳体或热管理支架要定料，把三样东西发过来就能给方向：长期水温、冷却液品类与浓度、装配预紧力或密封面公差。
料是同一个料，工艺是两套工艺——同一袋粒子发到两家厂，出来的一批能用十五年，一批两年就漏。
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。
节温器这类件的选料与试模，可以一起聊。
Last autumn, a client who makes thermal management modules sent two batches of thermostat housings.
Modified nylon parts, glass fiber reinforced, same grade, same sub-mold, same process card, from two shifts of the same factory.
One batch ran for eighteen months after being loaded, and the flange started to leak coolant; another batch has been almost three years, and when taken apart, it is still dry.
The customer's exact words were very direct:
"The material is the same material, and we also processed it the same way as them. Why did their batch crack, but ours didn't?"
I replied with three questions: What is the long-term water temperature? What is the concentration of ethylene glycol? What is the pre-tightening force of the flange bolts?
After asking three questions, the solution emerges. For the issue of thermostat housing materials, the answer is often not in the grade, but in the combination of the three numbers: 'temperature, medium, and preload'.
1. Six-dimensional working conditions: Here are two variables that others don't often calculate
The temperature should be considered over the long term. The thermostat housing is immersed in coolant at 105–130°C for extended periods, and the heat management module bracket on the side near the exhaust can be even higher.
The thermostat itself is a frequently moving part, with opening and closing cycles commonly on the scale of two to three minutes — it spends its days constantly alternating between hot and cold.
What does water at 130℃ mean? A household pressure cooker works at about 120℃. And this part has to be soaked in it continuously for tens of thousands of hours.
The medium is not 'ethylene glycol', it is 'coolant'. A typical mixture is 50% ethylene glycol and 50% water, with added corrosion inhibitors and defoamers.
Here is a key chemical change: ethylene glycol oxidizes under high temperature and aerobic conditions, producing small molecule acids such as glycolic acid and oxalic acid.
The result is that the coolant's pH will gradually drop from around 8 when new. When the pH falls to around 6, it becomes harsh on polyamides.
The pressure is fluctuating. The system pressure is commonly in the range of 0.1–0.2 MPa, varying with water temperature, and fluctuates multiple times throughout the day.
Among the loads, there is one that is 'static.' The preload of the flange bolts is applied on the housing over a long period, and what this brings is creep, not a strength issue.
The lifespan is calculated based on the whole vehicle: 15 years / 240,000 kilometers, during which the number of thermal cycles is in the tens of thousands.
Appearance and compliance. The flatness of the sealing surface, the smoothness of the internal flow channel, and the retention of performance after exposure to coolant—all three must be clarified in terms of standards before the fixed point.
Two number conversions:
First is the number of thermal cycles. Assuming two cold starts per day and 365 days a year, it amounts to around 700 times a year, and over fifteen years, it reaches an order of 10,000 times.
Secondly, creep. Plastic parts subjected to continuous pressure at 130°C often exhibit a deformation after 1000 hours that is an order of magnitude higher than the deformation under the same load at room temperature.
This second number is the real source of the flange leakage.
2. Material Path: Hydrolysis Resistance and Heat Resistance Are Not the Same
| Route | Long-term temperature resistance | Hydrolysis-resistant system | Water absorption magnitude | Cost |
|---|
| PA6-GF30 | 100–120℃ | Weak | approximately 2.8–3.0% | High risk of hydrolysis at 130°C for long periods |
| PA66-GF30 | 130–150℃ | Mature (compatible with thermally stable systems) | About 2.5% | Moisture-induced dimensional drift is relatively large |
| PA66-GF35 | 130–150℃ | Mature | About 2.2% | The strength of the weld line rises and falls with the glass fiber |
| PA6T-GF30 | Above 150℃ | Upper-middle | About 2%–3% | High cost, high mold temperature requirements |
| PPS-GF40 | 200℃ scale | Good | About 0.1% level | High cost, low toughness |
Look at this table, the key point is that 'heat resistance' and 'hydrolysis resistance' are two different things.
A material being able to withstand a short-term peak of 150°C does not mean it can soak in water and ethylene glycol at 130°C for three years — the latter tests the hydrolytic stability of the amide bonds.
The PA66 system is mainstream for this type of part, provided it is equipped with an appropriate heat stabilization and hydrolysis-resistant system; whether it is equipped or not, over the long term, the performance will diverge along the same line.
PA6T has a larger temperature margin, at the cost of process window and cost; PPS has good dimensional stability and chemical resistance, at the cost of toughness and cost.
The thermal management bracket and the thermostat housing often should not use the same part number: the former tends to favor structural rigidity, while the latter favors sealing and hydrolysis resistance.
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 |
|---|
| Mechanical retention after coolant exposure | After 130℃ × 1000h, tensile retention ≥ 70% | ISO 527-2 / GB/T 1040.2-2022 | Brittling, surface cracking | Hydrolysis-resistant system is configured according to temperature settings | Antioxidant (including thermally stable compounded system) |
| Molecular weight retention | The relative viscosity decreases within a controllable range after soaking | GB/T 1632.1 / ISO 307 | Long-term performance cliff-like decline | Moisture content control Hydrolysis-resistant system | Antioxidant (including thermally stable compounded system) |
| 130℃ Creep | 1000h deformation is determined per piece, focusing on the flange | Constant temperature and constant load creep test | Flange leakage, preload relaxation | Glass fiber reinforced Structural stiffeners | — |
| Maintain the flatness of the sealing surface | Still within drawing tolerances after thermal cycling | Coordinate Measuring Machine / Piece-Level Thermal Cycling | Leakage, improper sealing ring compression | Gate and Orientation Design | — |
| Inner flow channel precipitate | No visible precipitates, gating cross-section not reduced | Soak and then dissect, weigh | Valve port sticking, flow drift | Control the dosage of small-molecule additives | Lubricant (Low Bleed Type) |
| Weld line strength | ≥ 60% of the inherent strength | Component-level inspection Tensile | Weld line cracking | Control glass fiber content Gate design | Coupling Agent (Fiber / Resin Interface) |
| Blasting pressure | Follow item specifications, leave a margin | Hydraulic blasting test | Casing rupture | Structure and wall thickness coordination | — |
| Dimensional change after water absorption | within the magnitude of 0.3% | ISO 294 / Measured Before and After Humidity Adjustment | Assembly interference, sealing surface offset | Forced moisture conditioning or low water absorption substrate | — |
How to use this table: First look at the first two rows—mechanical retention and molecular weight retention after being resistant to coolant.
These two lines represent whether the material has been eaten away by water. If they don't pass, the creep and burst data afterward are only temporarily looking good.
A reminder: The coolant data must specify the type, concentration, pH, and temperature of the coolant. Data that only says 'ethylene glycol resistant' has limited reference value.
Four or Five Common Failures and Their True Causes
Failure 1: The flange surface slowly leaks, but the casing itself does not crack.
The root cause is mostly creep, not strength. When the bolt preload is applied on plastic at 130°C for a long time, the material will slowly flow away bit by bit.
Once the preload is released, the seal ring is not properly compressed, and leakage occurs. This cannot basically be detected in tests at room temperature.
Failure 2: Surface cracking and powdering, breaks with a simple twist.
This is hydrolysis plus long-term heat and oxygen. Why are water and acid so strong?
The main chain of polyamide is an amide bond, which can be cleaved by water and acid—molecular chains shorten segment by segment, and macroscopically, this results in strength collapse.
And the coolant becomes acidic after being used for a long time, which turns it from 'neutral water' into 'acidic water,' making it much more aggressive.
An attribution from the additive side: These components rely on thermal stability and hydrolysis-resistant systems to perform, but some compounded systems can be deactivated by complexation in coolant environments containing amines or sulfur. The same grade can have a significantly different lifespan in different coolants, and it's often due to this issue.
Failure three: Thermostat valve stuck, flow drift.
This is not a mechanical problem; it is precipitation. Small molecule additives are leached out by hot water and gradually accumulate near the valve port and valve seat.
During troubleshooting, first check the runner cross-section, then look at the amount of small molecules in the formula — this approach is more effective than changing the substrate.
Failure four: For the same batch of items, some can be used while others are defective.
First, check three things: dryness, assembly preload, and the brand and batch of the coolant. Any one of these being inconsistent is enough to create a difference in lifespan.
Failure Five: The corner near the exhaust side cracks first.
This is a typical signal of uneven temperature distribution. First, re-measure the temperature points on the whole machine, then discuss the materials.
Here's something that needs to be said directly: Many people think 'the more fiberglass is added, the more resistant it is to creep.'
The direction is half correct. Glass fiber can indeed increase the initial rigidity, but it cannot prevent the matrix from losing modulus under moisture absorption and high temperatures.
Moreover, after the glass fiber content increases, the interface between the fibers and the matrix can become the starting point for cracks in a long-term water environment, and the weld line strength also decreases accordingly.
So we are more inclined to ask: is this due to 'insufficient rigidity' or 'leakage at the sealing surface'? The solutions to these two issues are different.
5. Processing and Verification: Drying on this chain is the very first step
Dry. We ourselves have suffered losses in the workshop because of this: with the same batch of material, the same mold, and the parameters unchanged, one mold comes out fine, and the next mold breaks as soon as it's taken apart.
After checking the molds and parameters for three days, I couldn't find anything. Finally, I got to the dryer — those days a hot air dryer was being used.
Polyamide must be dried using a dehumidifying dryer; an ordinary hot air dryer is ineffective for it. Material with excessive moisture will hydrolyze and degrade in the barrel, and the strength of the parts has already been compromised.
Soaking in water at 130°C further shortens the lifespan. This kind of problem cannot be detected on the raw material bag; it only becomes apparent on the finished part, and it appears very late.
Mold temperature. The crystallinity is supported by the mold temperature. If the mold temperature is low, the parts become brittle, the surface darkens, and water resistance performance also decreases.
Gate and orientation. Near the sealing surface, it is desirable for the fiber orientation to be consistent to avoid local shrinkage differences that could warp the flatness.
It is recommended to arrange the verification sequence in this way:
1. Dimensions after humidity adjustment and flatness of the sealing surface (dry state data is only recorded for process documentation)
2. Mechanical retention and molecular weight changes after soaking in coolant at 130°C
3. 130℃ constant temperature and constant load creep (observe flange deformation)
4. Internal Pressure Circulation and Blasting
5. Sealing Test After Component-Level Thermal Cycling
6. Complete Vehicle Installation and Durability
The order cannot be changed. If the previous item fails, the numbers measured later will only look good temporarily.
An insider detail: the duration of a creep test cannot be shortened.
The deformation rate of the first three months cannot be extrapolated to the next three years—the creep curve of plastic is not a straight line; it can suddenly become steeper at a certain point in time.
6. Boundaries: When to return metals or switch systems
First, the long-term water temperature exceeds 140°C. In this range, the long-term performance of conventional polyamide systems is insufficiently supported, so one needs to move towards systems like PPS.
Secondly, if the coolant remains in the acidic range for a long time and cannot be replaced. If the coolant in the operating environment cannot be maintained periodically, the risk of hydrolysis will be amplified.
Third, it requires long-term zero leakage and no maintenance. The consequences of failure for this type of component are water leakage and shutdown, with a very high risk weight, so a structural compensation plan should be considered or it should revert to metal.
Fourth, the annual usage is too small to justify the cost of spreading the mold and long-cycle validation.
Laying out these four points is not to discourage, but to let the project know how long the validation will take during the project initiation stage.
Add a section: How to perform creep verification so that one dares to sign the report
This section is written for the person who has to sign. If the creep test is done incorrectly, it is more dangerous than not doing it at all because it will give you an overly optimistic number.
Temperature is taken according to long-term values. Using the creep curve at 23℃ to predict the deformation at 130℃ will definitely lead to an incorrect conclusion.
The load should be converted according to the actual preload, not just half. A conservative load leads to conservative conclusions, and vice versa.
Place the measurement point on the flange, not in the middle of the housing. What you need to answer is 'Will the sealing surface come loose?' not 'Is this piece of material good?'
All three time points should be recorded: 1000 hours, 3000 hours, and 5000 hours. A smooth initial period does not mean the later period will also be smooth.
At the same time, measure the pre-tightening force decay. The decay curve of the bolt torque is closer to the answer to the 'leakage-free' issue than the deformation value.
| Test item | Condition | What are you looking at? | Determine direction |
|---|
| constant temperature and constant load creep | Long-term water temperature, actual pre-tightening force | Flange deformation | Deformation determines the sealing margin |
| Preload force attenuation | Same as above | Bolt torque changes over time | Determining assembly torque and reinspection cycle |
| Mechanical properties after soaking | 130℃ × 1000h coolant | Stretch retention rate | Determine whether the material is eaten by water |
| Molecular weight change | Same as above | Relative viscosity | Degree of chain breakage |
| Component-level thermal cycling | -40℃ to 130℃ | Sealing surface flatness | Decide the sealed structure |
Adding another timeline — the same batch of material, two sets of processes, how the gap was created:
`
Same grade ├── Two shifts, two sets of drying and pre-tightening parameters, the first pieces are all qualified
│
Month 6
├── A small number of flange surfaces have white deposits (judged to be dirty coolant)
├── Occasional liquid level alarm after cold start (determined to be the sensor)
└── Month 18 concentrated report on seepage → Trace back → Ten to fifteen percent difference between the two batches of pretightening force → Retorque and drying window
`
The conclusion is not 'the material is no good.' The conclusion is that the same batch of material, under two different processes, has a lifespan difference of one and a half years.
This is also why, for the review of this type of part, I suggest including 'process and torque' in the technical agreement: whenever changing drying equipment or bolt suppliers, it must be reconfirmed each time.
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 sealing surface is compensated per piece. | Seal groove position and wall thickness change |
| Dry | Set the window according to the measured moisture content, using a dehumidifying dryer | Recycled materials mixed with the water content brought in |
| Material Temperature / Mold Temperature | Mold temperature is adjusted based on crystallinity and flatness jointly | Only copy the recommended brand numbers, without looking at the parts |
| Assembly Preload | Redetermine torque according to the creep characteristics of the material | Use the torque value of the original metal parts |
| Humidity control | Forced humidity adjustment Re-measure dimensions | Based on the average wall thickness to estimate the time, the thick-walled areas are not fully soaked. |
| Coolant Validation | Specify the category, concentration, pH, and temperature | Only make a group with pure ethylene glycol |
| Structure | The sealing groove and the stiffener need to be modified together. | Only change the materials, do not alter the structure |
| Verification order | Size → Medium → Creep → Internal Pressure → Complete Machine | If the previous item fails, just move on. |
One-page report sheet (for people who need to report upwards)
`
Project: Thermostat Housing / Thermal Management Module Bracket · Material Route Evaluation
Conclusion direction: Modified nylon can be considered as a candidate route, provided that both the hydrolysis-resistant system and structural compensation are in place
1. Three Rules That Must Be Followed
1. 130℃ coolant soaking data Molecular weight change data, both groups are required
2. The assembly preload should be recalibrated according to the creep characteristics and cannot follow the torque used for metal parts.
3. The duration of creep verification is not shortened, and no conclusion is drawn if the duration is not met.
2. Precondition (It is recommended to postpone if any are not met)
· Long-term water temperature within 140℃
· The coolant has a regular maintenance and replacement mechanism
· Creep test conditions with constant temperature and constant load
· The sealing surface can be controlled through structural compensation
3. Next Steps
1. Actual measurement of the coolant pH and temperature in the real usage environment
2. Soak at 130℃ × 1000h, and observe mechanical retention and molecular weight.
3. Perform constant-load creep at 130°C and measure the flange deformation
Risk warning: The main uncertainty of this route lies in long-term hydrolysis and creep, not in initial strength and temperature resistance.
`
Two questions readers often ask
Question: Compared with imported materials, where does the domestic route fall short?
According to public information, the advantage of imported brands in this type of component mainly lies in the integrity of long-term hydrolysis data, record of batch stability, and experience in validating supporting coolant systems.
The difference in domestic routes lies more in whether 'long-term data has been fully completed' rather than the materials themselves. Which components are already mature and which are still not recommended depends on the validation results of the components and cannot be generalized.
Q: Can the thermal management bracket directly use the part number of the thermostat housing?
Look at what it is afraid of. The bracket mainly fears rigidity and creep, while the casing mainly fears hydrolysis and sealing. The priorities of the two are different.
If the number of items is small and you want to combine part numbers for management purposes, you must at least re-verify creep under the condition of using the bracket — what is saved is management cost, not verification.
Conclusion
Returning to the three questions at the beginning. Why can these three questions set the direction?
Because it asks about three things: the water temperature (to determine the material system), the concentration (to determine the severity of the medium), and the preloading force (to determine the creep risk).
After these three questions are asked, it's finally the turn for the brand to appear.
If you currently have a thermostat housing or a thermal management bracket and need to decide on materials, just send over three things to give guidance: long-term water temperature, coolant type and concentration, and assembly preload or sealing surface tolerance.
The material is the same, but the processes are two different sets of processes — the same bag of particles sent to two factories, one batch can last fifteen years, while the other leaks in two years.
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 nylon resins, 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.
The material selection and test molding of parts like thermostats can be discussed together.
