去年有个做服务器风扇叶的客户，拿来一片改性尼龙注塑的扇叶。
叶片边缘有细微的缺料，他捏着叶尖转了转，说："这风扇用了半年怎么越来越响，是不是料强度不够？"
我们没先答强度，先问了转速和机房湿度。答案回来——五千转上下，常年不停，机房湿度不低。
这件事把服务器风扇选料的核心说透了：它的失效很少是"断"，多是"越转越响"。 响的根源在动平衡，不在强度。这一件要的从来不是"更结实"，是"转起来稳不稳"。
一、服务器风扇这一件，和汽车风扇不是一回事
先划清边界，免得套错结论。
汽车冷却风扇在发动机舱，长期 90–120℃，怕的是高温蠕变和疲劳，那是另一本账（汽车风扇篇单独讲过）。服务器风扇温度低得多——回路 40–60℃，风扇周边更不至于高到哪去。
但服务器风扇有自己的难：
转速高、常年不停。 服务器风扇常见每分钟数千到上万转，一年累计转动以千万次计，从不停机。汽车风扇随车速起伏，服务器风扇是 continuous duty。
动平衡极敏感。 叶片是高速旋转件，几克的重量分布变化，就会被放大成振动和噪音，再传给轴承和机箱。
洁净和阻燃是硬约束。 机房有人工作环境，要求低烟低毒，塑料件 UL94 V0 起步、多数无卤。
所以服务器风扇选料，考的不是"耐多高温度、扛多大载荷"，是"转十万小时后还平不平、稳不稳、干不脏"。
一句话：服务器风扇叶，强度是入场券，动平衡才是分水岭。
用量级分边界：汽车风扇周边 90–120℃，服务器风扇只有 40–60℃——温度差出一倍多，所以服务器风扇不用跟高温较劲，要把功夫花在转速和动平衡上。这也是它和汽车风扇不能用同一套料的原因。
二、工况六维：这件被什么约束
按六个维度摊开。
温度。 风扇周边 40–60℃，比汽车风扇温和。但长期连续，热老化是慢变量。
转速与载荷。 数千到上万转/分，离心力和风载叠加在薄壁叶片上。判据不是"断没断"，是"长期转动下重心偏不偏"。
介质。 不直接泡液，但有机房空气、凝露、灰尘。阻燃和低析出是重点。
寿命。 十年或数万小时连续运行。判据是动平衡保持率和噪音曲线，不是初始强度。
数万小时，按三万小时算，也相当于连续跑三年半不停。判据从"能不能转"变成"第三万小时还平不平整、响不响"。很多扇叶初始动平衡漂亮，慢漂半年就被投诉噪音——看寿命要看平衡寿命，不是强度寿命。
外观与洁净。 低烟低毒、低黄变、低析出，机房环境要求。
合规。 阻燃 UL94 V0，优先无卤；靠近电器看 GWIT。
六维里温度、转速、寿命都给了具体数字，件级精度就在这。
转速那行算笔账：五千转的扇叶，一年约转二十六亿圈；一万转的，约五十二亿圈。这么大的转动基数下，几毫克的重心偏移都会被放大成可感知的振动。动平衡不是"转得动"，是"转了一年还稳"。
三、叶片和风扇框，是两种料
这是"一套料走天下"方案最常踩的坑。
风扇叶片：要韧性（抗启动冲击）、要低翘曲（共面度）、要动平衡稳、要薄壁好充填。玻纤不能太高——GF15–25 是常见区间，太高各向异性放大、叶片反而脆。
为什么玻纤不能高？玻纤一高，流动时取向更明显，叶片不同位置收缩差更大，翘曲和根部应力都上来；薄壁长流程也难填，容易短射。GF15–25 是韧性和刚性、充填和翘曲的折中点，不是越低越好，也不是越高越稳。
风扇框：要刚性（固定电机和叶片）、要抗蠕变（长期预紧）、要尺寸稳、要阻燃。玻纤可以高一点，GF25–35 更合适。
| 要求 | 风扇叶 | 风扇框 |
|---|
| 首要性能 | 韧性 + 动平衡 + 低翘曲 | 刚性 + 抗蠕变 + 阻燃 |
| 玻纤含量 | GF15–25 | GF25–35 |
| 关键失效 | 动平衡劣化、根部疲劳 | 变形、装配松、阻燃析出 |
| 温度 | 40–60℃ | 与叶片接近 |
| 加工难点 | 薄壁充填 + 翘曲 + 平衡 | 收缩与尺寸 + 阻燃 |
同一个风扇总成，两个件两种配方。 用一套料，通常两边都不讨好。
为什么不能一套料？叶片要薄壁好充填、要低翘曲保共面，玻纤一高就脆、充填也难；框要刚性承电机、要抗蠕变，玻纤低了又撑不住。两件对玻纤含量的要求刚好错开，硬凑一套，只能两头妥协。
四、选型判据表：动平衡和尺寸稳定是主角
把约束落成可核对指标。下表门限是方向性建议，不是验收标准——实际由项目、工况和实测定。
| 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 |
|---|
| 动平衡（残余不平衡） | 按 ISO 1940 等级定，常 G6.3 内 | 动平衡机 + 全转速扫 | 振动、噪音、轴承磨 | 低翘曲 + 配重工艺 | 成核剂（尺寸稳定） |
| 弯曲模量（刚性） | 参考 6–10 GPa 量级 | ISO 178 | 叶尖变形、风量降 | 玻纤增强 | 偶联剂（界面增强） |
| 翘曲 / 共面度 | 叶片平面度按件定 | 三坐标 / 平台法 | 扫风、响 | 低翘曲体系 + 浇口 | 成核剂（结晶控制） |
| 吸湿重量 / 尺寸 | 湿态变化可控，不偏重心 | ISO 62 + 称量 | 动平衡漂移 | 低吸湿基材 | 材料本征，不靠助剂 |
| 长期耐温 | 60℃×长期 强度保持 | ISO 527 | 发白、脆化 | 热稳定体系 | 抗氧剂（热氧） |
| 阻燃 | UL94 V0，优先无卤 | UL94 / GWIT | 安规返工、析出 | 无卤阻燃体系 | — |
| 根部疲劳 | 长期转动无裂 | 耐久转动试验 | 叶片飞出 | 玻纤取向 + 浇口优化 | 偶联剂（界面） |
怎么用这张表：强度一行和动平衡一行分开打。只盯模量不盯动平衡，风扇转半年就响，强度再高也白搭。 带动平衡的件，吸湿尺寸那行权重很高。
一个实用提醒：带动平衡的件，把"湿态动平衡等级"单列红线，别和"干态模量"混在一张表里。混了，吸湿漂移被平均掉，出厂动静平衡都过，装机房湿一湿就破。
一个提醒：动平衡受吸湿影响这件事，常被漏掉。PA66 吸湿 8%，叶片吸湿后重量和尺寸都变，重心悄悄偏移——转速越高，这点偏移放大的离心力越大。 动平衡报告必须按调湿态出，不按干态。
五、四条常见误判，和真实根因
判反一：风扇响就是料强度不够。
这是最典型的误判。客户一听风扇响，第一反应是"料不行"。服务器风扇的响，多半来自动平衡劣化——吸湿让重心偏、翘曲让扫风、根部疲劳让裂。 这三件都和"强度够不够"不直接挂钩。我们接这类单，先问转速和机房湿度，再判断是动平衡还是疲劳，不急着换料。
判反二：常温脆断和低温脆断当一回事。
叶片测试里常出现"一次脆断"，但两件事处理方向完全不同。常温脆断多半是增韧不足、熔接线弱或干燥不够；低温脆断通常是增韧剂类型选错——普通弹性体在低温自己先脆。追问三句就能分：断在什么温度、先变形还是直接断、断面放射状还是拉丝。分不清，就一路换料一路不对。
判反三：用干态动平衡报告定设计。
尼龙吸湿会涨会重。PA66 吸湿 8%，叶片吸湿后重心偏移，动平衡等级掉档。用干态测的动平衡定装配，机房跑半年湿度上来，平衡就破了。 动平衡和尺寸都要按调湿态留余量。
判反四：阻燃加完就完事。
无卤阻燃加量大，某些阻燃剂或抗氧剂在长期热和加工高温下往表面迁移，出现析出、发雾、黄变。这是配方侧坑——加工温度超了耐温，或助剂耐温不够，表面就脏。 看到黄变析出，先查助剂耐温，别急着换基材。
一条时间线（行业常见的动平衡劣化路径）：风扇注塑、干态动平衡合格 → 装机房运行 → 半年吸湿、重心偏移 → 噪音上升、轴承磨损加剧 → 一年叶片扫风或根部微裂 → 追溯是吸湿尺寸没按湿态留余量。问题在选材阶段就埋了，只是慢。
六、加工与验证：动平衡和阻燃要分开盯
服务器风扇是注塑件，两个坑分开控。
干燥。 尼龙必烘，含水超标注塑降解，叶片根部强度掉、内应力大。干燥按实测含水率定窗口。
浇口与流道。 叶片薄壁长流程，玻纤取向决定根部强弱和翘曲。浇口位置先定好，不然后期救不回。
模温。 玻纤件模温提到 110–120℃ 量级，表面致密、浮纤少、翘曲小。
动平衡工序。 注塑后要做动平衡配重，且按湿态预期留余量。这一道工序省掉，后面全靠运气。
验证顺序。 建议这样排：
1. 材料级：弯曲模量、热氧保持率
2. 尺寸级：干湿态共面度、动平衡
3. 耐久级：长期转动 + 噪音曲线
4. 阻燃级：UL94 / GWIT
5. 系统级：上整机前最后验
顺序不能换。 前一项没过就往下走，后面数据没有解释意义。
七、边界：什么时候不该用改性尼龙
这一段可能比前面更值钱。
其一，超高转速或航空级风扇。 这类对动平衡和疲劳的要求超出热塑性常态区间，金属或特种材料更稳。
其二，长期温度超过 120℃ 的风扇位。 普通 PA66 体系不够，要半芳香族或 PPS 类。
其三，要求近零噪音极限的精密位。 尼龙的旋转噪声和吸湿波动，这类场景可能不接受，要换金属或特种。
其四，年用量小到摊不平注塑模和验证。 带动平衡的精密叶片要开模、做浇口优化、跑耐久，年用量几百个不成立。
其五，介质腐蚀或强静电环境。 这类要走专门体系，不是通用改性尼龙能兜底。
把这五条写在前头，不是劝退，是省时间。样品阶段平平稳稳、量产因动平衡或阻燃析出回退的项目，见过不止一个——回退代价比当初不做高得多。
八、换料风险清单（从原方案换到改性尼龙风扇叶 / 风扇框，要动什么）
| 环节 | 要动什么 | 容易漏的点 |
|---|
| 模具 | 叶片浇口按流动平衡定，不套旧位 | 玻纤取向导致根部弱 |
| 干燥 | 按实测含水率定窗口 | 回用料带入水分 |
| 料温 / 模温 | 玻纤件模温提到 110–120℃ | 只按牌号推荐值给 |
| 保压与脱模 | 根部与叶尖重点控 | 熔接线落高应力区 |
| 调湿 | 动平衡按湿态留余量 | 干态平、湿态偏 |
| 色差 | 外观件色板提前确认 | 阻燃体系黄变预期 |
| 验证顺序 | 材料 → 尺寸 → 耐久 → 阻燃 → 系统 | 前一项未过就往下走 |
九、一页纸汇报表（给要向上汇报的人）
`
项目：服务器风扇叶 / 风扇框 · 材料路线评估
结论方向：玻纤尼龙可作为候选，能否落地取决于三项前置
一、必须守住的三条
1. 动平衡按湿态定，不按干态
2. 叶片与风扇框两种配方，不一套料
3. 翘曲先动浇口，再谈换料
二、前置条件（任一不满足则建议暂缓）
· 长期工作温度 ≤ 60℃ 量级（局部另核）
· 有动平衡工序与湿态余量方案
· 年用量足以摊薄注塑与验证投入
三、下一步动作
1. 做干湿态共面度与动平衡
2. 做长期转动 + 噪音曲线
3. 定浇口与流道平衡方案
风险提示：本路线主要不确定性在动平衡与吸湿尺寸，不在初始强度。
`
十、读者常问的两句
问：和进口风扇尼龙差在哪？
只讲两件能对照的事：同一指标，看它标没标测试状态（干态还是湿态、全转速还是单点）；同一件上，看它给没给长期动平衡数据。风扇件对状态极敏感，状态不明的数字不宜直接比。有些叶片走国产 PA66-GF 增韧路线已经成熟，有些高平衡等级仍建议谨慎——具体到你的转速，要看温度和湿度两样。
问：PA66-GF 够不够，要不要上半芳香族？
看温度和位置。服务器风扇 40–60℃，PA66-GF 增韧体系够用也划算；只有长期超 120℃ 或贴热源的位置，才需要考虑半芳香族。 升级前先确认温升真到那档、结构余量已优化，再换不迟。多数服务器风扇，把钱花在动平衡和浇口上，比花在基材上升级更值当。
我们交付的，不只是一包料。
Last year, there was a customer who made server fan blades, and they brought a fan blade made of modified nylon injection molding.
The edges of the blades had slight material defects. He pinched the tip of a blade and turned it around, saying, 'This fan has been used for half a year, why is it getting louder and louder? Is it because the material isn't strong enough?'
We didn't ask about the intensity first; we asked about the rotation speed and the humidity in the machine room first. The answer came back — around five thousand revolutions per minute, running continuously all year, and the humidity in the machine room is not low.
This matter clearly explains the core of selecting materials for server fans: their failure is rarely a 'break', but more often 'becomes noisier as it spins.' The root of the noise lies in dynamic balance, not in strength. What is needed here has never been 'more solid', but 'whether it spins steadily'.
1. A server fan is not the same thing as a car fan.
First clarify the boundaries, so as not to draw the wrong conclusion.
Car cooling fans are in the engine compartment, lasting long at 90–120°C, and what they're afraid of is high-temperature creep and fatigue, which is another story (covered separately in the car fan section). Server fans are much cooler—the circuit is 40–60°C, and the area around the fan is not going to get very hot.
But the server fan has its own difficulties:
High rotation speed, running all year round. Server fans commonly rotate thousands to tens of thousands of times per minute, accumulating tens of millions of rotations in a year, never stopping. Car fans fluctuate with vehicle speed, while server fans are continuous duty.
Dynamic balance is extremely sensitive. The blades are high-speed rotating parts, and even a few grams of weight distribution change can be amplified into vibration and noise, which are then transmitted to the bearings and the casing.
Cleanliness and flame retardancy are hard constraints. The equipment room has a human working environment, requiring low smoke and low toxicity, with plastic parts starting at UL94 V0, mostly halogen-free.
So when selecting materials for server fans, it's not about 'how high a temperature they can withstand or how much load they can bear,' it's about 'whether they remain smooth, stable, and clean after ten thousand hours of operation.'
In a nutshell: For server fan blades, strength is the entry ticket, and dynamic balance is the watershed.
Boundary by magnitude: Car fan surroundings are 90–120℃, while server fans are only 40–60℃—the temperature difference is more than double, so server fans do not need to compete with high temperatures and should focus on speed and dynamic balance. This is also the reason why they cannot use the same materials as car fans.
2. Six-dimensional working condition: What constraints are applied to this part?
Spread out across six dimensions.
Temperature. Around the fan 40–60°C, milder than a car fan. But over long-term continuous use, thermal aging is a slow variable.
Rotational speed and load. Thousands to tens of thousands of revolutions per minute, with centrifugal force and wind load superimposed on thin-walled blades. The criterion is not 'whether it breaks,' but 'whether the center of gravity deviates under long-term rotation.'
Medium. Does not directly soak in liquid, but is exposed to organic room air, condensation, and dust. Flame retardancy and low outgassing are the focus.
Service life. Continuous operation for ten years or tens of thousands of hours. The criteria are the maintenance rate of dynamic balance and the noise curve, not the initial strength.
Tens of thousands of hours, taking 30,000 hours as an estimate, is equivalent to running continuously for three and a half years without stopping. The criterion has changed from "whether it can rotate" to "whether it remains smooth and quiet at the 30,000th hour." Many blades have a good initial dynamic balance, but after drifting for half a year, noise complaints arise — when looking at lifespan, we should consider balance lifespan, not strength lifespan.
Appearance and cleanliness. Low smoke and low toxicity, low yellowing, low precipitation, suitable for computer room environment requirements.
Compliance. Flame retardant UL94 V0, preferably halogen-free; check GWIT near electrical appliances.
In six dimensions, temperature, rotational speed, and lifespan are all given specific numbers; the component-level precision lies here.
Let's do a little calculation about the rotation speed: a fan blade spinning at 5,000 RPM makes about 2.6 billion revolutions in a year; at 10,000 RPM, it's about 5.2 billion revolutions. With such a large number of rotations, even a few milligrams of imbalance in the center of mass can be amplified into perceptible vibrations. Dynamic balancing is not 'spinning smoothly,' it's 'still stable after spinning for a year.'
3. The blades and the fan frame are made of two different materials.
This is the most common pitfall of the 'one set of materials to go around the world' plan.
Fan blades: They need toughness (resistant to startup impact), low warpage (coplanarity), good dynamic balance stability, and thin walls for good filling. The glass fiber content cannot be too high—GF15–25 is the common range; too high and anisotropy increases, making the blades brittle.
Why can't the glass fiber content be high? When the glass fiber content is high, the orientation becomes more apparent during flow, leading to greater shrinkage differences at different positions of the blade, increasing warpage and root stress; thin-walled, long-flow parts are also difficult to fill and prone to short shots. GF15–25 represents a compromise between toughness and rigidity, as well as filling and warpage, and it's not the case that lower is always better or higher is always more stable.
Fan frame: it should be rigid (to fix the motor and blades), creep-resistant (long-term preloading), dimensionally stable, and flame-retardant. The glass fiber content can be a bit higher, GF25–35 is more suitable.
| Requirement | Fan blade | Fan frame |
|---|
| Primary performance | Toughness Dynamic Balance Low Warping | Rigidity Creep Resistance Flame Retardant |
| Glass fiber content | GF15–25 | GF25–35 |
| Critical failure | Dynamic balance deterioration, root fatigue | Deformation, loose assembly, flame retardant precipitation |
| Temperature | 40–60℃ | Close to the blade |
| Processing difficulties | Thin-wall filling Warping Balance | Shrinkage and Size Flame Retardant |
The same fan assembly, two parts, two formulas. Using one set of material usually doesn't satisfy either side.
Why can't the same material be used? The blades need to have thin walls for good filling and low warping to maintain coplanarity. If the glass fiber content is too high, they become brittle and difficult to fill; the frame, on the other hand, needs rigidity to support the motor and resist creep, and if the glass fiber content is too low, it won't hold up. The two parts have conflicting requirements for glass fiber content, so forcing a single material would only lead to compromises at both ends.
4. Selection Criteria Table: Dynamic Balance and Dimensional Stability are the Main Focus
Turn constraints into verifiable indicators. The thresholds in the table below are directional suggestions, not acceptance standards — the actual values are determined by the project, working conditions, and measurements.
| Indicator | Directional Threshold | Verification Method / Standard | Common Failures | Common solution | Corresponding auxiliary agent system |
|---|
| Dynamic balancing (residual imbalance) | According to ISO 1940 classification, usually within G6.3 | Dynamic balancing machine: full speed sweeping | Vibration, noise, bearing wear | Low warping Counterweight technology | Nucleating agent (size-stable) |
| Bending modulus (rigidity) | Refer to the 6–10 GPa range | ISO 178 | Blade tip deformation, reduced airflow | Glass fiber reinforced | Coupling agent (interface enhancement) |
| Warp / Coplanarity | Blade flatness is determined per piece | Coordinate Measuring Machine / Platform Method | Sweeping wind, sound | Low warp system Gate | Nucleating agent (crystallization control) |
| Moisture Absorption Weight / Size | Wet-state changes are controllable and the center of gravity is not biased. | ISO 62 Weighing | Dynamic balance drift | Low moisture-absorbing substrate | Intrinsic material properties, without relying on additives |
| Long-term heat resistance | 60℃ × Long-term Strength Retention | ISO 527 | Pale and brittle | Thermally stable system | Antioxidant (thermal-oxidative) |
| Flame retardant | UL94 V0, halogen-free preferred | UL94 / GWIT | Safety standard rework, precipitation | Halogen-free flame retardant system | — |
| Root fatigue | Long-term rotation without cracks | Durability rotation test | The leaf flies out | Glass fiber orientation Gate optimization | Coupling agent (interface) |
How to use this table: record the row for strength and the row for dynamic balance separately. Focus only on the modulus, not on dynamic balance. The fan will start making noise after half a year, and no matter how high the strength is, it will be useless. For parts that involve dynamic balance, the moisture absorption dimension row carries a very high weight.
A practical reminder: For parts requiring dynamic balancing, list the 'wet dynamic balancing grade' separately in red, and don’t mix it with the 'dry modulus' in the same table. If mixed, moisture-induced drift gets averaged out, both dynamic and static balancing pass at the factory, but once installed in a humid machine room, it will fail.
A reminder: The fact that dynamic balance is affected by moisture absorption is often overlooked. PA66 absorbs 8% moisture, and after the blades absorb moisture, both their weight and dimensions change, subtly shifting the center of gravity—the higher the rotational speed, the greater the centrifugal force amplifying this slight shift. Dynamic balance reports must be based on the conditioned state, not the dry state.
5. Four Common Misjudgments and the Real Causes
Judgment one: The fan noise indicates that the material strength is insufficient.
This is the most typical misjudgment. When the customer hears the fans running, their first reaction is 'the material is no good.' The noise from server fans is most often due to imbalance deterioration—moisture absorption causing off-center weight, warping affecting airflow, and root fatigue causing cracks. None of these three are directly related to 'whether the strength is sufficient.' When we handle such cases, we first ask about the fan speed and the humidity in the server room, then determine whether it is imbalance or fatigue, without rushing to replace the material.
Judgment 2: Room temperature brittle fracture and low temperature brittle fracture are considered the same thing.
In blade testing, "brittle fracture" often occurs, but the handling approaches for the two situations are completely different. Brittle fracture at room temperature is mostly due to insufficient toughening, weak weld lines, or inadequate drying; brittle fracture at low temperature is usually because the wrong type of toughening agent was chosen—ordinary elastomers become brittle by themselves at low temperatures. You can distinguish them with three quick questions: at what temperature did it break, did it deform first or break directly, and is the fracture cross-section radial or fibrous. If you can't tell, you'll just keep changing materials without success.
Counter-triple three: Use the dry-state dynamic balance report to determine the design.
Nylon absorbs moisture, which causes it to swell and become heavier. PA66 absorbs 8% moisture, and after the blades absorb moisture, the center of gravity shifts, causing the dynamic balance grade to drop. Using dynamic balance measured in the dry state for assembly, after running in the machine room for half a year, as humidity rises, the balance will be ruined. Both dynamic balance and dimensions must allow for tolerance according to the conditioned moisture state.
Judgment 4: Once the flame retardant is added, it's done.
Halogen-free flame retardants need to be added in large amounts. Some flame retardants or antioxidants can migrate to the surface under long-term heat and high processing temperatures, resulting in exudation, fogging, or yellowing. This is a formulation-side issue — if the processing temperature exceeds the heat resistance or the auxiliary agents are not heat-resistant enough, the surface will become dirty. When you see yellowing and exudation, first check the heat resistance of the additives, and don't rush to change the substrate.
A timeline (industry common path of dynamic balance deterioration): fan injection molding, dry-state dynamic balancing qualified → installed and running in the machine room → half a year of moisture absorption, center of gravity shift → noise increase, accelerated bearing wear → after one year, blade wind sweep or slight root cracks → the trace is that moisture absorption dimensions did not leave allowance for wet state. The problem was buried in the material selection stage, it just happens slowly.
6. Processing and Verification: Dynamic balancing and flame retardancy should be monitored separately
The server fan is an injection-molded part, with the two slots controlled separately.
Drying. Nylon must be baked; if the moisture content exceeds the standard, molding will degrade, causing reduced strength at the base of the blades and high internal stress. Drying should be based on the measured moisture content to determine the window.
Gates and runners. The blades have thin walls and long flow paths, and the orientation of the glass fibers determines the strength and warpage at the root. The gate position must be set first, otherwise it cannot be recovered later.
Mold temperature. Raising the mold temperature of fiberglass parts to the range of 110–120°C results in a dense surface, fewer floating fibers, and minimal warping.
Dynamic balancing process. After injection molding, dynamic balancing weights need to be added, leaving a margin according to the expected wet state. If this step is skipped, everything afterward depends on luck.
Verification order. It is recommended to arrange it like this:
1. Material level: flexural modulus, thermo-oxidative retention rate
2. Dimensional level: coplanarity in dry and wet states, dynamic balance
3. Durability Level: Long-term Rotation Noise Curve
4. Flame Retardant Grade: UL94 / GWIT
5. System level: Final inspection before assembling the whole machine
The order cannot be changed. If the previous item hasn't passed, moving on will make the subsequent data meaningless.
7. Boundaries: When not to use modified nylon
This section might be more valuable than the previous one.
Firstly, ultra-high-speed or aviation-grade fans. Such fans have requirements for dynamic balance and fatigue that exceed the normal range of thermoplastics, so metals or special materials are more stable.
Secondly, fan positions where the long-term temperature exceeds 120°C. Ordinary PA66 systems are not enough; semi-aromatic or PPS types are required.
Third, precision positions requiring near-zero noise limits. The rotational noise of nylon and moisture absorption fluctuations may not be acceptable in such scenarios; metal or special materials need to be used instead.
Fourth, the annual usage is too small to justify the cost of injection molding and validation. Precision blades that require balancing need tooling, gate optimization, and durability testing, which is not feasible with an annual usage of only a few hundred units.
Fifth, corrosive media or strong static environments. These require specialized systems; general modified nylon cannot cover them.
Writing these five points at the beginning is not to discourage, but to save time. Projects that go smoothly during the sample phase but are later rolled back during mass production due to dynamic balance or flame retardant precipitation—I've seen more than one; the cost of rollback is much higher than not doing it in the first place.
8. Material Change Risk Checklist (What needs to be changed when switching from the original plan to modified nylon fan blades / fan frame)
| link; segment; part | What needs to be moved? | Points that are easy to overlook |
|---|
| Mold | The blade gate is determined according to flow balance, without using the old position | Fiber orientation causes weak roots |
| Dry | Set the window according to the measured moisture content | Moisture carried in by recycled materials |
| Material Temperature / Mold Temperature | The mold temperature of the fiberglass parts is raised to 110–120°C | Give only according to the recommended value by grade |
| Pressure holding and demolding | Focus on controlling the roots and leaf tips | Splice line high stress zone |
| Humidity control | Dynamic balancing is based on the wet-state allowance | Normal when dry, slightly abnormal when wet |
| Color difference | Advance confirmation of exterior color samples | Expected yellowing of flame-retardant system |
| Verification order | Material → Size → Durability → Flame Retardant → System | Proceed if the previous item is not passed |
9. One-page report form (for those reporting upward)
'
Project: Server fan blades / fan frame · Material route evaluation
Conclusion: fiberglass nylon can be a candidate, but whether it can be implemented depends on three prerequisites
1. The three essential points to be preserved
1. Dynamic balance is determined by wet state, not dry state
2. Two formulations: blade and fan frame, not a single material
3. Warping should be moved to the gate first, then material replacement should be discussed
2. Prerequisites (if any one is not met, postponement is recommended)
· Long-term operating temperature ≤ 60°C (local details to be separately)
· Dynamic balancing process and wet margin plan with
· Annual usage is sufficient to dilute injection molding and verification investment
3. Next steps
1. Perform wet and dry coplanar coplanarity and dynamic balance
2. Perform long-term rotation noise curve
3. Gate and runner balance solution
Risk warning: The main uncertainty in this route lies in dynamic balance and moisture absorption size, not in initial strength.
'
10. Two common questions readers often ask
Question: What's the difference compared to imported nylon fans?
Let's just talk about two things you can compare: for the same indicator, check if it marks the test state (dry or wet, full speed or single point); For the same item, see if it provides long-term dynamic balance data. Fan components are extremely sensitive to condition, so it's best not to compare numbers with unknown status. Some blades are already mature in the domestic PA66-GF toughening route, but for some high balance grades, caution is still advised—when it comes to your speed, you need to consider temperature and humidity.
Question: Is PA66-GF sufficient? Should I use the upper half of the aromatic mode?
Check temperature and location. Server fans are 40–60°C, PA66-GF toughening systems are sufficient and cost-effective; Only long-term temperatures above 120°C or locations near heat sources should semi-aromatic elements be considered. Before upgrading, confirm the temperature rise is true and that structural allowance is optimized; then replacement is not too late. Most server fans are more worthwhile to spend on dynamic balancing and gates than upgrading to base materials.
What we deliver is more than just a package of materials
