Guides / Shop Practice

What to Do When a Large Machined Part Fails After Shipping

October 6, 2026

A quality escape on a small part is a bad day. On a large part it is a bad month, because the part that failed weighs several tonnes, took weeks to make and cannot simply be swapped out of a bin for a spare.

That size problem changes the response. Remaking the part from scratch is rarely realistic on the timeline the buyer needs, and pretending otherwise wastes the time that should go into finding a fix that actually works.

Why remaking is usually the wrong first answer

A new blank, if it is a casting or a large forging, can take as long to arrive as the original part took to build in full. Add the machining hours back on top and a straight remake often means eight to twelve weeks the buyer’s schedule does not have.

The buyer’s actual goal is rarely “an identical new part”. It is a working assembly on their timeline. That opens up options a straight remake forgets to consider:

• Rework in place, if the deviation can be corrected by removing more material or adjusting a mating feature within the design intent.
• A field fix, such as line boring a bore back to size or building up and re-machining a worn or out-of-tolerance face, sometimes on site rather than shipping the part back.
• An engineered deviation, agreed with the buyer’s engineering, if the part still performs its function outside the original tolerance and they are willing to accept it, documented and signed off, not assumed.
• A genuine remake, kept as the fallback when none of the above meets the function, with an honest lead time attached from the start.

Propose the realistic option first, with the reasoning behind it, rather than defaulting to the most expensive answer out of caution.

Telling the buyer without wrecking the relationship

The instinct to gather every fact before saying anything is understandable and usually costs more trust than it protects. Buyers judge a shop on how a problem is handled at least as much as on whether the problem happened.

Report as soon as the finding is confirmed, even if root cause is still open. “We have identified an out-of-tolerance condition on the bore, we are still confirming the cause, here is what we know so far” is a message a buyer can plan around. Silence for a week while the investigation runs is not.

State the finding plainly, with the actual measurement, not a softened version of it. Buyers on large parts are engineers. A vague description reads as evasive even when it is not intended that way.

Bring a proposed path, not just the problem. “Here is what we found, here is what we think caused it, here is what we propose to do and by when” moves the conversation from blame toward resolution in one message.

Own what is yours to own. If the cause sits with the shop, say so directly. Buyers who work with heavy machining long enough have seen escapes before. What damages the relationship is a shop that argues the finding rather than one that owns it and fixes it.

Root cause and containment, in writing

A buyer evaluating the response will expect a written account covering the same ground a formal 8D or corrective action report covers, whether or not that exact format is used.

1. The finding. What was measured, where, against what requirement.
2. Immediate containment. What happens to this part right now, and whether any other parts from the same run or blank need checking.
3. Root cause. Not “operator error” as a final answer, but the actual mechanism: a fixture that shifted, a tool that wore past its limit undetected, a datum scheme that allowed stack-up nobody caught.
4. Corrective action. What changes in the process so the same escape cannot recur, not just on this part but on the next one like it.
5. Verification. How the fix on this specific part will be proven, and by what inspection method, before it ships or is returned to service.

Skipping straight from finding to fix, without the root cause step written down, tends to produce a repeat escape later, because the actual mechanism was never identified.

Why both sets of records matter

The buyer’s own inspection records are usually what caught the problem in the first place, which is one more reason a shop should want a buyer who inspects carefully rather than one who does not. The shop’s own inspection records, from the original job, are what show whether the escape was a measurement gap on the shop’s side or a condition that developed after the part left, such as damage in transit or a change during installation.

Both records get compared during the root cause step, and a shop with thin records of its own has a much harder time showing where the failure actually happened. Keeping full dimensional and process records on every large part, not only the ones that end up in dispute, is what makes that comparison possible when it is needed.

What to do next

Write the corrective action template now, before the next escape, covering finding, containment, root cause and verification in one page. Having the structure ready before it is needed is the difference between a fast, credible response and a scramble that costs the relationship along with the part.

Related reading: quality plans and inspection test plans for large parts, inspecting large machined parts and building trust with a new buyer without testimonials.

Frequently asked questions

Keep reading

• Running a Heat Treat and NDT Supply Chain, Not Just a Machine
• Planning Safety and Site Access Before the Crew Travels