电话打进来的时候是晚上八点多。
一个做水泵的客户，开口一句："我们那个牌号，厂家说今年都不排了。"
他用了六年的一款进口 PA66-GF30，突然被告知这个规格要停。库存还剩不到二十天。他当天已经问了四家供应商，其中三家都说"有相近的"，有一家甚至说"一模一样"。他问我该选哪个。
我先问了另一个问题：手上这批库存，按现在的用量还能撑多少天？
他愣了一下。这个问题听起来跟"找替代"没关系，但它决定了后面所有的路：他剩下的时间，决定了他是可以从容地从第一步走到第四步，还是**必须在两天内赌一个"。
答案是二十天。二十天意味着够他做完前面两步，不够他从最省事的那一步开始。
后来他采取的办法是：先用现有库存把最后一批关键件打出来，同时并行启动"同牌号换来源"和"相近牌号验证"两条线，其中相近牌号那条提前跑工艺窗口测试。
断供这件事，往往不是慢慢发生的。
可能是一条产线检修、一次船期延误、一纸出口政策，或者某个牌号突然限量——供应说紧就紧。
这时候最忌讳的动作是"看到有货就买"。替代要按顺序做，顺序错了，比断供本身更麻烦。
一、先判断：是断一阵，还是断定了
| 类型 | 典型特征 | 应对重点 |
|---|
| 短期波动 | 交期拉长、限量供货 | 保主力牌号，压缩非关键件 |
| 中期紧张 | 价格上行、渠道收紧 | 提前备关键批号 |
| 长期断供 | 停产、退出、无明确恢复期 | 启动正式替代流程 |
判断依据不是听消息，是看三样东西：原厂通告、渠道实际到货、连续几周的交期变化。
这三样的重要程度是递进的。通告是意图，渠道到货是现实，而连续几周的交期变化是趋势——趋势最难造假。一家说"没问题"的供应商，如果连续三周把你说好的交期往后挪，那它嘴上的"没问题"就没有任何信息量。
只听到"好像要涨"就大批囤货，和确认了停产还在等消息，是两种同样贵的错误。
还有一个更实际的判断依据：直接问原厂要"恢复时间"。
愿意给出明确时间点的，通常属于短期波动；含糊其辞、或者干脆说"暂无计划"的，就要按长期断供来准备。这条信息比任何渠道传话都准——因为它是供应方自己的排产判断，不是外界的猜测。
问的时候有个小技巧：不要问"还能不能供货"，要问"下一批什么时候能排"。 前者会得到一句含糊的安慰，后者会得到一个具体的日期或者一个具体的推脱。推脱本身就是答案。
另外有一类情况要单独识别：规格还在，渠道断了。 有时候原厂并没有停产，只是某个区域的代理体系调整了，导致你原来的那条路走不通。这种情况下急着做替代验证其实是白做——换个有授权的渠道就能解决。判断缺料之前，先判断是不是渠道有问题。
二、替代的四步顺序（不能跳）
1. 先确认能不能等效替代——指标对照，找出关键指标（不是全部指标）
2. 同牌号换来源——同等级换渠道，最快、风险最低
3. 同体系相近牌号——同基材、相近含量，工艺微调
4. 换体系——比如 PA66 换 PA46 / PA6T 之类，成本与工艺都要重估
这个顺序是从"最省事"到"最费事"排列的。
很多人直接跳到第四步，理由是"只看到这个有货"。但第四步的代价是：工艺要重调、模具可能要改、认证可能要重走。能用第一步解决的，绝不该走到第四步。
把每一级的代价摆出来，这个顺序为什么不能跳就很清楚了：
| 替代层级 | 通常要重做的事 | 大致周期 | 隐性代价 |
|---|
| 第 1-2 级：等效 / 换源 | 指标对照、小批验证 | 一到两周 | 几乎为零 |
| 第 3 级：相近牌号 | 工艺窗口重调、性能验证 | 三到六周 | 调机时间、一批试错料 |
| 第 4 级：换体系 | 模具评估、认证重走、长期老化验证 | 两到六个月 | 可能改模，认证费重交 |
这张表里最不直观的是第 3 级和第 4 级之间的落差。这两个级别在上游距离很近（都是换一支料），但在下游距离很远——因为第 4 级跨过了"体系"这条线，而跨体系意味着收缩率、结晶行为、成型窗口全都变了。
举个具体的：从 PA66 换到 PA46，收缩率的变化可能让你的公差链整个重算。 一个原本刚好合格的孔径，换料之后可能系统性偏小；而这不是批次问题，是这套模具配这套料的结构性问题。解决它的方法是改模具，而不是调参数。
所以真正该问的不是"这个能不能替代"，而是"我跨到第几级了"。跨到第 3 级的时候要开始算调机的钱；跨到第 4 级的时候要先确认模具和认证愿不愿意跟着动。
三、替代能不能成，看三张表
物性对照表——关键指标的差异，逐项列出
工艺窗口对照表——料温、模温、干燥条件、收缩率
认证文件对照表——UL、安规、客户认可清单
第三张表最容易漏，也最容易让前期全部白费。
一个件性能完全对得上，但文件链接不上客户的认证体系，项目照样走不动。所以替代方案评审时，三张表要一起看，不能先看性能、后补文件。
第二张表里最该盯的是收缩率，因为它是唯一一项能单独把"合格"变成"不合格"的东西。
物性上的差异，很多时候在这张表里并不显眼——拉伸强度差 5% 通常不影响使用；但收缩率差 0.2 到 0.3 个百分点，在精密件上就是从"公差内"推到"公差外"。而这个后果不会在小批量验证里暴露，它会等到你连续生产几千件的时候，以尺寸超差的形式一次性出现。
所以做工艺窗口对照的时候，收缩率这一栏不要只看数值，要拿它去推一遍你现有图纸上的关键公差。 这个动作花不了半小时，但它能在改模之前把问题拦下来。
四、替代的三个禁区
1. 认证件不做等效声明就直接换。 客户认可要提前走流程，不能换完再说。
这一条最常见的结果是：料换了，件做出来了，性能没问题，然后在客户审厂或者年度复核的时候被翻出来。这时候要么承认私自变更——这是合规问题；要么把已经生产的产品全部追溯——这是灾难。 两种都不好玩，而避免它只需要提前走一次变更申请。
2. 只看拉伸强度。 尼龙真正的关键常在耐温、吸湿、尺寸稳定、长期老化——拉伸强度往往是最容易对上、也最不说明问题的一项。
这一条要解释一句为什么：因为拉伸强度是最容易通过配方调整补回来的一项。玻纤多加几个点、换更高黏度的基材，拉伸强度就上去了。但耐温、吸湿、长期老化这几项是材料本征属性决定的，改配方很难弥补。所以拿最容易对上的指标去做匹配，等于在最没有区分度的地方验证——结果当然都对，下了线却慢慢出问题。
3. 换料不换工艺。 尤其换到高温尼龙，料温、模温、干燥要求完全不同，工艺不换等于白换。
高温尼龙这一项要特别提：PA46、PA6T、PA9T 这一族的加工温度普遍比 PA66 高出一截，模具温度也常常要上到一百多度。如果你的模温机上不去，或者料筒加热圈功率不够，这支料在你的机台上就是打不好的——这时候你会觉得"这个料不行"，其实是设备跟不上。换体系之前先看一遍设备能力，这一步省不得。
五、替代方案定下来之后，要跑的验证
替代方案定了，不等于能用。至少跑三项验证：
一是短期性能对照。 关键指标做同条件对比测试，看差异是否在可接受范围。
二是工艺窗口测试。 新牌号的料温、模温、干燥条件要重新确认，不要沿用原牌号的参数。
三是加速老化对比。 尤其对耐久件，短期性能一样不等于长期一样。热老化、湿热老化之后的指标保留率差异，往往要到使用半年才暴露。
这三项里，第三项最容易跳过，代价也最大。 因为它跳过的不是一次测试，是一整段使用周期。
加速老化这个方法本身值得说一句：它不是为了得到精确答案，是为了制造一个可比较的差异。把新旧两种料放在同一个严苛条件下跑一段时间，看谁的指标掉得快——这个相对关系，比任何绝对值都能说明长期表现。哪怕具体时长对应不上实际使用年限，只要两条曲线的斜率不同，结论就已经出来了。
六、怎么把断供风险提前降下来
关键牌号保持"双来源"，平时就有一家备用
备一段安全库存，覆盖的是交期，不是囤货
提前了解可替代方向，不要等断供当天才开始找
和供应商建立变化通报机制，让信息先于断供到达
最好的替代方案，是在没断供的时候就已经验证过的那个。
这句话反过来也成立：断供当天才想到去做替代验证的项目，注定要付一次停线成本。
"双来源"这一条要补一个细节：备用的那一家不是留个联系方式就行，它要真的下过单、真的验过货。因为名录上的第二来源和实际能供货的第二来源，中间隔着一次完整的采购流程。很多企业在断供那天才发现，自己"有"的那个备用供应商，上一次合作已经是三年前了。
安全库存那条也要说清：它覆盖的是"从发现断供到找到可用替代"的周期，不是"永远不断货"。 这个周期如果包括走到第 3 级替代（三到六周），那安全库存就要备到六周以上；如果行业里这批件最多能容忍两周断档，那就要想办法把验证流程压缩到两周以内。先算这个周期，再定库存数量，否则要么不够用，要么压一堆用不上。
七、把这件事变成制度，而不是应急
建议每年做一次"关键牌号体检"：把所有在用牌号列一张表，每一项标注三栏——有没有第二来源、有没有已验证的替代方案、安全库存能覆盖多少天。
三栏里有空的，就是下次断供时的风险点。这张表不需要多复杂，一页纸就够。但它的作用是把"断供焦虑"变成"可管理的清单"——能提前看到的缺口，都不算真正的缺口。 这三栏每年更新一次，就是一份自己的供应风险地图。
这张表还有第二个用处：它能告诉你该先补哪一栏。 一个牌号如果有第二来源但没有验证过替代方案，你要做的是抽空跑一轮验证；如果有替代方案但没有第二来源，你要做的是找一家能长期供的同等的。两种缺口的补救方式完全不同，而在没列表之前，它们看起来是一样的焦虑。
结语
回到开头那个客户。
他的最后一个库存撑了二十六天，比预期的二十天多出六天——多出来的那部分，是因为他在断供消息确认的第一天就停掉了非关键件的生产，把料全部留给关键件。这个动作让他多出了做第 2 级验证的时间。
最终他没有走到第 4 级。同牌号换了一家有授权的渠道，加上一支相近牌号做备份——前者解决当下，后者解决下一次。整个过程没有停线。
替代不是"换个能买的"，是"换个能用的"。
顺序对了，替代是计划；顺序错了，替代是赌博。
如果你正卡在某个牌号上，把原牌号和用途发来，我们帮你从第一步开始对。
The phone rang after eight o'clock in the evening.
A customer who makes water pumps said at the beginning, 'Our grade, the manufacturer said they won't produce it this year.'
He had been using an imported PA66-GF30 for six years, and was suddenly told that this specification would be discontinued. His inventory had less than twenty days left. He had already asked four suppliers that day; three of them said they had 'similar ones,' and one even said 'exactly the same.' He asked me which one to choose.
I first asked another question: With the current usage, how many more days can this batch of stock last?
He was momentarily stunned. This question didn't seem related to 'finding a substitute,' but it determined all the steps that followed: the time he had left decided whether he could calmly move from the first step to the fourth, or if he **had to gamble on it within two days.
The answer is twenty days. Twenty days means enough for him to complete the first two steps, but not enough for him to start from the easiest step.
Later, the method he adopted was: first use the existing inventory to produce the last batch of key components, while simultaneously initiating two parallel lines—'switching sources with the same grade' and 'validation of similar grades,' with the latter testing the process window in advance.
The act of cutting off supply often doesn't happen gradually.
It could be a production line maintenance, a shipping schedule delay, an export policy, or suddenly a certain grade being limited — supply can become tight in an instant.
At this time, the most taboo action is 'buying as soon as you see stock.' Instead, you should follow the steps in order; if the order is wrong, it is more troublesome than a supply cutoff itself.
1. First, determine: is it a temporary break, or a final decision?
| Type | Typical features | Key response |
|---|
| Short-term fluctuation | Extended delivery time, limited supply | Preserve the main brand, compress non-critical components |
| Mid-term tension | Price increase, channel tightening | Prepare key batch numbers in advance |
| Long-term payment interruptions | Discontinued, withdrawn, no clear recovery period | Initiate the official replacement process |
The basis for judgment is not hearsay, but looking at three things: the official manufacturer announcement, the actual arrivals through channels, and the changes in delivery times over several consecutive weeks.
The importance of these three things is progressive. A notice represents intention, the arrival of goods through the channel represents reality, and changes in delivery dates over several consecutive weeks represent a trend — trends are the hardest to fake. If a supplier says 'no problem,' but pushes back the agreed delivery date for three consecutive weeks, their verbal 'no problem' carries no information at all.
Hearing 'it seems like the price will go up' and rushing to stockpile, and waiting for news even after confirming production has stopped, are two equally costly mistakes.
There is also a more practical criterion: directly ask the manufacturer for the 'recovery time'.
Those willing to give a specific time point usually reflect short-term fluctuations; those who are vague or simply say 'no plans yet' should be prepared for a long-term supply cut. This information is more reliable than anything heard through other channels—because it is the supplier's own production schedule judgment, not external speculation.
There's a small trick when asking: don't ask 'Can you still supply it?' Instead, ask 'When can the next batch be scheduled?' The former will get you a vague reassurance, while the latter will get you a specific date or a specific excuse. The excuse itself is the answer.
There is another situation that needs to be identified separately: the specifications still exist, but the channel is cut off. Sometimes the original manufacturer hasn't discontinued the product; it's just that the agent system in a certain region has been adjusted, making your original route unavailable. In this case, rushing to do alternative validation is actually pointless — switching to an authorized channel can solve the problem. Before judging a shortage of materials, first determine whether the issue is with the channel.
2. The alternative four-step sequence (cannot be skipped)
1. First, confirm whether it can be equivalently replaced — compare the indicators and identify the key indicators (not all indicators).
2. Change source for the same brand — change channel for the same grade, fastest and lowest risk
3. Similar grades within the same system — same base material, similar content, slight process adjustments
4. Change the system—for example, switching from PA66 to PA46 / PA6T, both cost and process need to be reassessed
This order is arranged from 'the easiest' to 'the most troublesome'.
Many people jump directly to the fourth step, reasoning that 'I only see this is in stock.' But the cost of the fourth step is: the process has to be readjusted, the mold may need to be changed, and certification may have to be redone. If it can be solved in the first step, you should never proceed to the fourth step.
Lay out the cost at each level, and it becomes very clear why this sequence cannot be skipped:
| Alternative hierarchy | Things that usually need to be redone | Approximate cycle | Hidden cost |
|---|
| Level 1-2: Equivalent / Source Switching | Indicator comparison, small batch validation | One to two weeks | Almost zero |
| Level 3: Similar Grades | Process window readjustment and performance verification | Three to six weeks | Machine setup time, a batch of trial and error material |
| Level 4: Change System | Mold evaluation, recertification, long-term aging verification | Two to six months | May need to modify the model, certification fee needs to be paid again |
The least intuitive part of this table is the gap between level 3 and level 4. These two levels are very close upstream (both involve switching a material), but they are far apart downstream—because level 4 crosses the 'system' line, and crossing systems means the shrinkage rate, crystallization behavior, and molding window all change.
To give a specific example: switching from PA66 to PA46, the change in shrinkage rate could require recalculating your entire tolerance chain. A hole that was originally just within specification might systematically become smaller after changing the material; and this is not a batch issue, but a structural issue of this mold paired with this material. The way to solve it is to modify the mold, not adjust parameters.
So the question we should really ask is not 'Can this replace it?' but 'Which level have I reached?' When reaching level 3, it is necessary to start calculating the setup costs; when reaching level 4, we first need to confirm whether the mold and certification are willing to move along.
3. Whether alternatives can succeed, look at three tables
Physical Properties Comparison Table — Differences in Key Indicators, Listed Item by Item
Process Window Comparison Table — Material Temperature, Mold Temperature, Drying Conditions, Shrinkage Rate
Certification Document Comparison Table — UL, Safety Standards, Customer Approval List
The third form is the easiest to overlook and also the easiest to make all previous efforts go to waste.
Even if a component's performance fully meets the requirements, if the documentation cannot connect with the customer's certification system, the project still cannot move forward. Therefore, during the alternative solution review, the three forms must be examined together; you cannot check the performance first and add the documentation later.
The item that should be monitored most closely in the second table is the shrinkage rate, because it is the only factor that can alone change 'qualified' to 'unqualified'.
Differences in material properties are often not obvious in this table— a 5% difference in tensile strength usually does not affect usage; but a 0.2 to 0.3 percentage point difference in shrinkage rate can mean going from 'within tolerance' to 'out of tolerance' for precision parts. And this consequence will not show up in small batch tests; it will only appear all at once in the form of dimensional deviations when you produce thousands of pieces consecutively.
So when doing process window comparison, don't just look at the shrinkage rate numbers—use it to calculate the key tolerances on your existing drawings. This process takes less than half an hour, but it can block the problem before mold changes.
4. The three forbidden zones for substitution
1. Replace certified parts without an equivalent declaration. Customer approval must go through the process early; don't wait until after the replacement.
The most common result is: the material is changed, the part is made, performance is fine, and then it is revealed during the customer's factory audit or annual review. At this point, either admit unauthorized changes—this is a compliance issue; Or trace all the products already produced — which is a disaster. Neither is fun, and to avoid this, you only need to submit a change request in advance.
2. Only look at tensile strength. The real key to nylon is often its temperature resistance, moisture absorption, dimensional stability, and long-term aging—tensile strength is often the easiest to match and least speaks to.
This point explains why: because tensile strength is the easiest to compensate for through formulation adjustments. Adding a few more points to fiberglass and switching to a higher viscosity substrate increases tensile strength. But temperature resistance, moisture absorption, and long-term aging are determined by the material's intrinsic properties, and changing the formula is hard to compensate for. So using the easiest indicators to match is like verifying in the least distinguishable areas—the results are all correct, but problems gradually arise once the product is offline.
3. Change materials but not processes. Especially when switching to high-temperature nylon, the material temperature, mold temperature, and drying requirements are completely different; not changing the process is basically replacing for nothing.
High-temperature nylon deserves special mention: PA46, PA6T, PA9T families generally have processing temperatures higher than PA66, and mold temperatures often exceed 100 degrees. If your mold temperature controller can't get up, or if the barrel heating ring power is insufficient, this material won't work well on your machine—at that point, you'll feel "this material isn't good," but it's actually because the equipment can't keep up. Before switching systems, check the equipment's capabilities—this step can't be skipped.
5. After the alternative plan is decided, verify what needs to be done
Having an alternative plan doesn't mean it's usable. At least run three validations:
First, short-term performance comparison. Perform comparative tests under the same conditions for key indicators to see if the differences are within acceptable ranges.
Second, process window testing. The material temperature, mold temperature, and drying conditions of the new grade must be re-confirmed; do not use the parameters of the original grade.
Third, accelerate aging comparison. Especially for durable parts, short-term performance is not the same long-term. Differences in retention rates after thermal aging and damp heat aging often only become apparent after six months of use.
Among these three items, the third is the easiest to skip and the most costly. Because it is skipped not just one test, but an entire usage cycle.
The accelerated aging method itself is worth mentioning: it is not about obtaining precise answers, but about creating a comparable difference. Put both new and old materials under the same harsh conditions for a period of time to see whose indicators drop faster—this relative relationship better than any absolute value indicates long-term performance. Even if the exact duration doesn't correspond to actual usage years, as long as the slopes of the two curves differ, the conclusion is already set.
VI. How to Lower Supply Interruption Risk in Advance
Key Grades Maintain "dual sources". Usually, there's a backup
with a safety stock segment covering delivery times, not stockpiling
to understand alternative directions in advance. Don't wait until the day of supply interruption to start contacting
to establish a change notification mechanism with suppliers. The best alternative to get information before the supply interruption reaches
is the one that has already been verified before the supply interruption.
's statement holds the opposite: if you only think to do a substitution verification project on the day of the supply cutoff, you will inevitably have to pay the cost of a shutdown.
The "dual source" section needs to be detailed: the backup supplier isn't just about providing contact information; it must have actually placed an order and inspected the goods. Because the second source listed on the directory and the actual supply source are separated by a complete procurement process. Many companies only realize on the day of supply cutoff that their "available" backup supplier last cooperated three years ago.
Safety stock also needs to be clarified: it covers the cycle "from discovery to supply interruption to finding available replacement," not "never out of stock." If this cycle includes reaching Level 3 replacement (three to six weeks), then safety stock must be kept for more than six weeks; If the industry can tolerate a two-week gap in this batch, then find a way to compress the verification process to under two weeks. First, calculate the cycle, then set the inventory quantity; otherwise, either it's insufficient or you can't use a pile of stock.
7. Make this a system, not an emergency
suggests doing a "key grade check-up" once a year: list all the brands in use in a single table, each marked with three columns—whether there is a second source, if there are verified alternatives, and how many days the safety stock can cover.
The blank sections in the three columns are the risk points for the next supply interruption. This table doesn't need to be complicated; one page is enough. But its purpose is to turn "payment interruption anxiety" into a "manageable list"—any gaps you can see in advance are not real gaps. These three columns are updated once a year, serving as your own supply risk map.
This table has a second use: it tells you which columns to fill in first. If a grade has a second source but hasn't verified an alternative, you need to make time to run a round of verification; If there is an alternative but no second source, you need to find a company that can supply the same quality long-term. The two gaps have completely different ways to remedy them, and before listings, they look like the same anxiety.
Conclusion
Back to the customer at the beginning.
His last inventory lasted twenty-six days, six days more than the expected twenty—the extra part was because he stopped production of non-critical parts on the first day the supply cut was confirmed, leaving all materials for the critical parts. This move gave him extra time for Level 2 verification.
In the end, he didn't reach Level 4. If you switch to an authorized channel with the same grade, add a similar brand as a backup—the former solves the present, the latter handles the next time. The whole process is offline.
Substitute isn't 'buy something you can buy', it's 'switch to something usable.'
If the order is right, substitute is a plan; If the order is wrong, substitute is gambling.
If you're stuck on a certain grade, send us the original number and purpose, and we'll help you start from step one
