What If a Correction Shipment Fixes Most of Your 3D Printed Parts but a Few Are Still Bad?

April 19, 2026

A correction shipment can feel like the job is almost done, but mixed results are where buyers often lose control. Maybe most of the new parts fit, yet two still crack at install. Maybe the revised latch works on nine units, but one batch still binds on the mating feature. That does not mean the whole job failed. It means you now have to separate what is proven from what still needs attention.

The expensive mistake is treating a mostly successful correction shipment like full approval. The better move is to sort the parts clearly, show what changed, and keep the next decision tied to evidence instead of relief.

Fast route:

Use the close-but-still-wrong guide if you are still dealing with the first sample.
Use the receiving check guide if the correction shipment just landed and you need a clean intake process.
Request the next quote step when you have sorted the good units from the ones that still miss.
Use the reorder-without-the-original guide if the corrected job is mostly proven and you need a stable repeat baseline later.

First separate good, questionable, and failed units

Do not leave the whole shipment in one pile while you argue about whether it is "basically fixed." Mixed outcomes need physical separation first.

Good: installs cleanly, functions correctly, and matches the corrected expectation.
Questionable: may fit, but needs force, shows inconsistent finish, or behaves differently from the successful units.
Failed: clearly misses a fit point, cracks, binds, warps, or cannot be trusted in service.

If you mix those groups together, the next conversation becomes muddy fast. Buyers end up describing averages instead of real pass/fail behavior.

What this result usually means

A mixed correction shipment often means one of three things happened:

the correction solved the main geometry issue but did not fully remove variation from a sensitive feature
the original problem description was directionally right, but one secondary issue stayed hidden until the remake was tested at quantity
the design is now close enough that acceptance depends on a tighter definition of what counts as good

That is why a mixed result should trigger sorting, photos, and unit counts instead of broad emotional summaries like "most are fine" or "a few still feel weird."

Run a controlled receiving check instead of ad hoc bench testing

When the correction shipment arrives, use the same check sequence on each unit so the pass/fail line stays consistent:

label or count every unit before testing
test the same fit points in the same order
separate good, questionable, and failed units immediately
photograph at least one passing example and one failing example
record the exact count in each group

If you need a fuller intake flow, use the receiving inspection guide before the evidence gets scattered.

The most useful evidence to send back after a mixed correction shipment

Evidence	Why it helps
Pass/fail counts	Shows whether the problem is isolated, recurring, or still systemic.
One successful installed photo	Proves what correct fit now looks like.
One failed installed photo	Lets the shop compare the miss against the passing version.
Marked close-up of the mismatch	Keeps the next correction tied to the real contact point instead of a vague complaint.
Any pattern by position or subgroup	Helps detect whether only certain units or conditions still trigger the miss.

Do not let the good units erase the bad ones

Buyers sometimes talk themselves into accepting the whole shipment because the correction clearly improved things. Improvement matters, but a few bad units still need to be named and counted. Otherwise the next reorder starts from a false story: "the remake fixed it."

A more useful summary sounds like this: The correction fixed the clip alignment on most units, but 3 of 18 still bind at the lower guide rail and should not be treated as approved.

When it is reasonable to keep the good units and continue

Keeping the good units can make sense when:

the pass/fail line is easy to define
the good units are genuinely serviceable and consistent
the failed units show a clear repeat issue that can be isolated
everyone agrees the shipment is not the same thing as full batch approval

That lets the buyer move forward with what is proven while still documenting what remains unresolved.

When you should pause reorders or larger quantities

Pause the next quantity if any of these are true:

you still cannot explain why some units pass and others fail
the questionable units only work with force, trimming, or selective assembly tricks
the remaining miss affects a safety, retention, sealing, or load-bearing feature
the corrected part still depends on loose language like "most seem okay"

At that point the job is not ready for reorder logic yet. It is still in controlled correction mode. If you need that broader downstream process, keep the reorder-control guide in view.

A clear message template for mixed-good correction results

We tested all 18 corrected units. 15 install and function correctly. 3 still fail at the lower guide rail and require force that the passing units do not need. Attached are one installed passing photo, one installed failing photo, and marked close-ups of the remaining interference point. Please quote the next correction step based on the 3 failing units only, not as full production approval.

That keeps the next step readable for both sides.

How this connects to the wider buyer path

This page sits between the first-sample correction loop, receiving control, and reorder consistency. The main goal is simple: do not let a partly successful correction shipment blur the line between evidence and approval.

Common questions

Should I reject the entire shipment if only a few corrected parts are still bad?
Not automatically. First sort the units clearly and document the pass/fail counts. The next decision depends on how defined the remaining issue is.

Can I keep the good units and still request help on the bad ones?
Yes, if the good units are truly proven and the remaining failed units are documented as a separate unresolved group.

What if the questionable units only work when installed carefully?
Treat them as questionable or failed, not good. If the part needs special handling to hide the problem, it is not yet a clean pass.

Does a mostly successful correction shipment mean the next reorder is safe?
No. Reorders should wait until the remaining misses are either resolved or explicitly accepted as non-critical.

What to open next after a partly successful correction shipment

Use the next page that fits the remaining unresolved risk

Still in fit-correction mode

Go back to the close-but-still-wrong loop
Use this when the remaining miss still needs a sharper geometry correction before you treat anything as stable.

Need clean remake language

Sort the pass/fail split clearly
Use this when the next challenge is separating proven units from unresolved ones without blurring approval.

Thinking about future orders already

Hold off and lock reorder consistency first
Use this only after the remaining misses are resolved clearly enough to define the real accepted baseline.

If the correction loop keeps dragging, tighten the control points around it

Reorder baseline

Open Asset 09
Use this when a partly fixed sample is getting mistaken for a safe production baseline.

Correction-cost reset

Open Asset 25
Use this when the remaining miss means the next step is another scoped change, not a vague free extra try.

Release discipline

Open Asset 26
Use this when the bigger problem is that a partly fixed correction result is being treated like full release approval too early.

If the remake batch may also be mixing old and new versions, read What Should You Do If a Custom 3D Printing Shipment Includes Mixed Revisions in the Same Batch?.

Move from this article into the next useful decision.

Use the main hubs to troubleshoot visible defects, compare materials, tighten print settings, check relevant reviews, or figure out whether the work should be quoted or produced professionally.