The Repair LibraryRead · Learn · Master

Destructive vs. Non-Destructive Analysis

A failure analysis lives on evidence, and this section is the toolkit for gathering it — organized around the single distinction that governs how the whole toolkit is used: whether a method preserves the evidence or consumes it. Every analytical technique falls into one of two families. Non-destructive analysis examines a part while leaving it intact — visual inspection and microscopy, X-ray that sees inside a package without opening it, thermal imaging that finds a hot fault, electrical characterization that measures behavior — and its defining virtue is that the part survives, so the examination can be repeated, checked, and built upon, and the evidence remains available for whatever the analysis needs next. Destructive analysis, by contrast, consumes the part to reveal what non-destructive methods cannot reach — decapsulation that removes a chip's package to expose the die, cross-sectioning that cuts and polishes a part to show its internal structure — and it is powerful precisely because it goes where nothing else can, but it is irreversible: once a part is decapped or cross-sectioned, the original is gone and cannot be un-cut. From that irreversibility follows the one rule that governs the entire toolkit, non-destructive first, always: because a destroyed part cannot be re-examined, every non-destructive method that could answer the question is exhausted before any destructive one is begun, and the analysis climbs a ladder from the least invasive method to the most, going destructive only at the end and only when the answer genuinely requires seeing inside and is worth the part. The section teaches that ladder, and the discipline that comes with it: that a destructive step, when it is finally justified, is aimed by the non-destructive findings rather than taken blind — the X-ray that located the defect telling the cross-section exactly where to cut — because a destructive step gets one attempt and a blind one often destroys the evidence without revealing it. It also treats the real hazards of the destructive methods honestly, the aggressive acids of chemical decapsulation and the cutting and grinding of cross-sectioning, which demand proper facilities and protection and are often best sent to a lab rather than attempted casually. The point it leaves is that most repair-level failure analysis never needs to destroy anything at all — the non-destructive toolkit answers the question — and the professional discipline is to know the whole ladder, to climb it in order, and to reserve the irreversible step for the rare case where the internal answer is both required and worth the part it costs.

ProfessionalMedium Risk23 min read

What You Will Learn

  • You will learn the distinction between non-destructive analysis, which preserves the part, and destructive analysis, which consumes it.
  • You will learn the governing rule — non-destructive first, always — and why a destroyed part cannot be re-examined.
  • You will learn the escalation ladder from least invasive to most, and where destructive methods sit at its end.
  • You will learn to aim a destructive step with the non-destructive findings, because it gets one attempt.
  • You will learn when a destructive step is justified — when the internal answer is required and worth the part — and the hazards it carries.

What You Will Be Able To Do

  • You will be able to classify an analytical method as non-destructive or destructive and know which preserves the evidence.
  • You will be able to apply the non-destructive-first rule and exhaust the preserving methods before any consuming one.
  • You will be able to climb the escalation ladder from least to most invasive in the right order.
  • You will be able to use non-destructive findings to aim a destructive step rather than take it blind.
  • You will be able to judge when a destructive step is justified, and recognize the hazards that make it a lab's task.

Required Tools

  • A microscope, X-ray access, and a thermal camera — the non-destructive methods that are exhausted first, in escalating order
  • Electrical test gear — non-destructive characterization of a part's behavior before anything invasive
  • The non-destructive findings — used to aim any destructive step, so a cut lands where the evidence is
  • A qualified lab's contact — because chemical decapsulation and cross-sectioning are hazardous and often best outsourced

When NOT to Attempt This

Do not attempt this section if any of the following apply to you:

  • You are not experienced with the specific repair type described here.
  • You do not have professional-grade equipment for this procedure.
  • The device has sentimental or high monetary value and you cannot afford a mistake.
  • You have not successfully completed this repair on a sacrificial device first.

Section Overview

A failure analysis lives on evidence, and this is the toolkit for gathering it, organized around one distinction: does a method preserve the evidence or consume it (failure-analysis-purpose-and-process). Every technique falls into two families. Non-destructive analysis examines a part while leaving it intact — visual and microscopy, X-ray that sees inside without opening, thermal imaging that finds a hot fault, electrical characterization — so the part survives and the examination can be repeated (x-ray-inspection-when-its-needed). Destructive analysis consumes the part to reveal what non-destructive methods cannot reach — decapsulation that exposes the die, cross-sectioning that shows internal structure — and it is irreversible (thermal-cameras-hotspot-detection-in-repairs). From that follows the governing rule: non-destructive first, always. A destroyed part cannot be re-examined, so every non-destructive method is exhausted before any destructive one, climbing a ladder from least invasive to most (root-cause-analysis-techniques). And a destructive step, when justified, is aimed by the non-destructive findings — the X-ray locating the defect telling the cut where to land — because it gets one attempt. Most repair analysis never destroys anything; the discipline is to know the ladder, climb it in order, and reserve the irreversible step for when the internal answer is required and worth the part.

Why This Matters

This is the discipline that keeps a failure analysis from destroying the very evidence it needs (failure-analysis-purpose-and-process). This matters because irreversibility is absolute: a part decapped or cross-sectioned is gone, and with it any chance to re-examine it, run a different method, or recover from a mistake — so the order of operations is not a preference but a one-way door, and going through it too early forfeits everything behind it (x-ray-inspection-when-its-needed). This matters because the non-destructive toolkit is more capable than it seems: X-ray sees inside a sealed package, thermal imaging finds a fault by its heat, electrical characterization reads a part's behavior — so most questions can be answered without destroying anything, and reaching for a destructive method first is usually reaching past the answer (thermal-cameras-hotspot-detection-in-repairs). It matters because a destructive step gets one attempt: there is no second cut, so the step must be aimed by what the non-destructive methods found, and a blind cross-section through a part often misses the defect and destroys the evidence in the same stroke (root-cause-analysis-techniques). And it matters because the hazards are real: chemical decapsulation uses aggressive acids and cross-sectioning uses cutting and grinding, so these are not casual bench steps but hazardous procedures needing proper facilities, and knowing that is part of knowing when to send a part to a lab rather than attempt it. Exhaust the non-destructive methods, climb the ladder in order, aim any destructive step, and respect its hazards — and the toolkit answers the question without needlessly destroying the evidence or the technician.

Required Prerequisites

Before starting this section, you should have completed:

  • Failure Analysis — Purpose and Process — the analysis that needs evidence and the discipline of preserving it; this section is the toolkit for gathering that evidence.
  • Root Cause Analysis Techniques — the reasoning that each analytical method supplies evidence for; a destructive step is justified only when the root cause genuinely requires the internal view.
  • A method-ladder card — the non-destructive methods in escalating order, then the destructive ones, kept as a list so the analysis climbs the ladder rather than jumping to the bottom.
  • The full non-destructive record — the images, X-rays, and thermal and electrical readings, kept complete, because they are both the answer to most questions and the aim for any destructive step.
  • Proper protection for destructive work — acid-rated protection, fume handling, and eye protection are prerequisites for any chemical or mechanical destructive step, and their absence is itself a reason to outsource.
  • A failed part answerable non-destructively — a fault visible by X-ray, thermal, or electrical means, so the non-destructive toolkit is practiced to a real answer without destroying anything.
  • X-ray, thermal, and microscopy access — the escalating non-destructive methods, to practice climbing the ladder in order.
  • A scrap part with a lab-obtained internal view — ideally the die exposed through a qualified lab, so the destructive step is understood without a casual attempt at hazardous chemistry.

Real-World Applications

This distinction shapes how a technician gathers the evidence a failure analysis needs. A bench facing a failed IC reaches first for X-ray and thermal imaging, answering most questions without opening the package (x-ray-inspection-when-its-needed). A technician who has exhausted the non-destructive methods and still needs the internal view plans the decapsulation, or sends the part to a lab, only because the answer genuinely requires seeing the die (failure-analysis-purpose-and-process). A bench preparing a cross-section uses the X-ray that located the defect to aim the cut, so the one attempt lands where the evidence is (thermal-cameras-hotspot-detection-in-repairs). And a tech weighing a destructive step judges whether the internal answer is worth the part and within safe reach, or whether the analysis is complete enough already (root-cause-analysis-techniques). The confusions this prevents: a part destroyed before the non-destructive methods were tried, a destructive step taken blind and missing the defect, a hazardous decapsulation attempted without facilities, and a part consumed for an answer the non-destructive toolkit already held.

Common Challenges

  • Irreversibility is easy to underestimate. A destructive step feels like just another method until it is done — and then the part is gone, with no second attempt and no recovery (x-ray-inspection-when-its-needed).
  • The non-destructive toolkit is underused. The reach of X-ray, thermal, and electrical methods is easy to miss — and reaching for a destructive method first often reaches past an answer already available (thermal-cameras-hotspot-detection-in-repairs).
  • A blind destructive step wastes the part. A cut taken without knowing where the defect is — often destroys the evidence without revealing it, spending the one attempt on nothing (root-cause-analysis-techniques).
  • The hazards are real and easily mishandled. Chemical decapsulation uses aggressive acids and cross-sectioning uses cutting and grinding — casual attempts risk injury, and their danger is part of the decision to outsource (failure-analysis-purpose-and-process).

Safety Notes

Risk Level: Medium. The non-destructive methods are low-risk, but the destructive ones carry real chemical and mechanical hazards, so this section's safety framing is emphatic and the standing bench law applies throughout.

  • Chemical die exposure is a lab procedure — fuming acids demand a fume hood, acid-rated protection, and proper disposal; for most technicians the part is sent to a qualified lab, not decapped on the bench.
  • Cross-sectioning has mechanical hazards — cutting, grinding, and polishing create sharp edges and hazardous dust and require proper equipment and protection.
  • Non-destructive first is also the safe default — the low-risk methods answer most questions, so the hazardous destructive step is reached rarely and reluctantly.

Professional Tips Before Starting

  • Classify the method first. Preserve or consume — know which family a method is in before using it, because that decides where it sits on the ladder (x-ray-inspection-when-its-needed).
  • Exhaust the non-destructive methods. X-ray, thermal, electrical, all before any cut — most questions are answered without destroying anything (thermal-cameras-hotspot-detection-in-repairs).
  • Aim any destructive step. Use the non-destructive findings to place the cut — the one attempt must land where the evidence is (root-cause-analysis-techniques).
  • Weigh the part against the answer. A destructive step spends the part — take it only when the internal answer is required and worth the cost (failure-analysis-purpose-and-process).
  • Respect the hazards or outsource. Acids and cutting are dangerous — without proper facilities, the part goes to a lab, not the bench.

The Two Families and the Ladder

Non-Destructive and Destructive — The Two Families

The whole toolkit of failure analysis sorts into two families, and which family a method belongs to decides everything about how it is used (failure-analysis-purpose-and-process). The first family is non-destructive analysis. These are the methods that examine a part while leaving it whole: visual inspection and microscopy that read a surface, X-ray that sees through a package to the structure inside without opening it — non-destructive for essentially all repair work, though a high cumulative X-ray dose can disturb the contents of flash and other floating-gate memory, so a data-bearing part is not dosed carelessly, thermal imaging that finds a fault by the heat it makes, acoustic microscopy that finds internal voids and delamination, and electrical characterization that measures how the part behaves — and their shared, defining virtue is that the part survives (x-ray-inspection-when-its-needed). Because the part survives, a non-destructive examination can be repeated, cross-checked, and built upon: a reading questioned can be taken again, a second method can be tried on the same part, and the evidence remains available for whatever the analysis needs next. The second family is destructive analysis. These are the methods that consume the part to reach what non-destructive methods cannot: decapsulation, which removes a chip's protective package — chemically with acids or mechanically — to expose the silicon die for direct examination; cross-sectioning, which mounts, cuts, grinds, and polishes a part to reveal its internal structure in a plane; and delidding and similar operations that open a part up. Their power is that they go where nothing else can — inside the package, into the die, through the layers — and their cost is that they are irreversible: a part decapped or cross-sectioned is gone, its original state unrecoverable, and there is no second copy of that specific failure to examine again (thermal-cameras-hotspot-detection-in-repairs). Preserve or consume, survive or spend — the two families, and the distinction the whole rest of the section turns on.

Non-Destructive First — The Governing Rule and the Ladder

From the irreversibility of the destructive family comes the one rule that governs the entire toolkit: non-destructive first, always (x-ray-inspection-when-its-needed). The reasoning is simple and absolute. A destroyed part cannot be re-examined, so any question that a non-destructive method could answer must be put to that method first, while the part still exists to answer it — because if a destructive step is taken first and the non-destructive answer is later needed, the part to provide it is already gone. This turns the analysis into a climb up a ladder. The methods are ordered from least invasive to most, and the analysis climbs from the bottom: visual and magnified inspection first, then X-ray and thermal and electrical characterization, exhausting the non-destructive rungs — and only when those are genuinely exhausted, and the answer still requires seeing inside, does the analysis step onto the destructive rungs at all (thermal-cameras-hotspot-detection-in-repairs). And even among the destructive methods the order holds. The less consuming destructive step comes before the more consuming one — a decapsulation that exposes the die is often tried before a cross-section that cuts the part in two, because it leaves more of the part for a later step (root-cause-analysis-techniques). The ladder is not bureaucracy but preservation. Each rung is climbed only when the ones below it have been exhausted, because every rung upward forecloses something, and the destructive rungs foreclose the part itself — so a technician never jumps to the bottom of the ladder when a higher rung would have answered the question. Least invasive first, most invasive last, and destructive only at the end — the ladder that keeps the analysis from destroying its own evidence.

Aiming the Cut and Weighing the Part

When a destructive step is finally justified, two disciplines govern it: it is aimed, not blind, and it is weighed, not automatic (root-cause-analysis-techniques). A destructive step gets exactly one attempt. There is no second cut of the same part, so a cross-section taken without knowing where the defect lies is a gamble that usually loses — the plane of the cut misses the failure, and the part is consumed with nothing to show (thermal-cameras-hotspot-detection-in-repairs). So the destructive step is aimed by the non-destructive findings. The X-ray that located a void, the thermal image that pinpointed a hot spot, the electrical test that isolated a shorted region — each tells the destructive step exactly where to go, so that the one cut lands on the defect and the one decapsulation exposes the right region, which is why the non-destructive record is not discarded when the destructive step begins but is precisely what guides it (x-ray-inspection-when-its-needed). The second discipline is to weigh the part against the answer. A destructive step spends the part permanently, so it is worth taking only when the internal answer is genuinely required — when the root cause cannot be established without seeing inside — and when that answer is worth more than the part it costs, which for a repair is often not the case at all: a repair frequently needs only enough of the cause to fix the device, not a full internal post-mortem. So most repair-level analysis stops before the destructive rungs entirely. The non-destructive toolkit answers the question, the part is preserved, and the destructive step is reserved for the rare case where the internal view is both necessary and justified (failure-analysis-purpose-and-process). Aim the cut with what you found, and spend the part only when the answer is worth it — the two disciplines that make a destructive step, when it comes, deliberate rather than wasteful.

Common Mistakes

  • Destroying the part before the non-destructive methods were tried. A destructive step is taken while a non-destructive one would have answered — the part is gone, and with it the chance the preserving method offered (x-ray-inspection-when-its-needed).
  • Taking the destructive step blind. A cut or decap is made without knowing where the defect is — the one attempt misses, and the evidence is destroyed without being revealed (thermal-cameras-hotspot-detection-in-repairs).
  • Jumping to the bottom of the ladder. The most invasive method is reached for first — the ladder is climbed from the least invasive, and skipping rungs forecloses evidence a lower rung would have preserved (root-cause-analysis-techniques).
  • Spending the part for an answer already in hand. A part is consumed for an internal view the non-destructive record already provided — the destructive step adds nothing and costs the part (failure-analysis-purpose-and-process).
  • Attempting hazardous destructive work casually. Acid decapsulation or cross-sectioning is improvised without facilities — the hazards are real, and the right course is often a qualified lab, not the bench.

Troubleshooting Guidance

  • You are tempted to open a part to see inside — try the non-destructive methods first: X-ray, thermal, and electrical characterization answer most questions without destroying anything, so exhaust them before any cut, because the part cannot be un-opened (x-ray-inspection-when-its-needed).
  • A destructive step is genuinely needed — aim it with the non-destructive findings: use the X-ray or thermal result that located the defect to place the cut or the decap, so the one attempt lands on the evidence rather than missing it (thermal-cameras-hotspot-detection-in-repairs).
  • You are unsure whether to go destructive at all — weigh the part against the answer: a destructive step is justified only when the internal view is required to establish the cause and worth the part it spends, which for a repair is often not the case (root-cause-analysis-techniques).
  • The destructive step involves acids or cutting you cannot do safely — outsource it: chemical decapsulation and cross-sectioning are hazardous and need proper facilities, so a part that requires them and cannot be handled safely goes to a qualified lab (failure-analysis-purpose-and-process).

Verification & Testing Methods

Confirm your grasp of the toolkit before the chapter's report:

  • [ ] I can classify a method as non-destructive analysis or destructive analysis and say which preserves the evidence.
  • [ ] I can apply the non-destructive-first rule and exhaust the preserving methods before any consuming one.
  • [ ] I can climb the escalation ladder from least to most invasive, with decapsulation and cross-sectioning at its end.
  • [ ] I can aim a destructive step with the non-destructive findings rather than take it blind.
  • [ ] I can judge when a destructive step is justified and recognize the hazards that make it a lab's task.

Then try the practice exercises below — non-destructive escalation and destructive-decision reasoning only; no hazardous destructive work is performed.

Practice Exercises

  1. Classify the methods (6 minutes, desk work). For a list of analytical methods — visual, X-ray, thermal, electrical, decapsulation, cross-sectioning — label each non-destructive or destructive and order them from least to most invasive, building the ladder (failure-analysis-purpose-and-process).
  2. Climb the non-destructive ladder (7 minutes, a part with a findable fault). On a failed part, work up the non-destructive rungs — visual, then magnified, then X-ray and thermal or electrical — until the fault is found, answering the question without destroying anything (x-ray-inspection-when-its-needed).
  3. Aim a destructive step on paper (7 minutes, from the non-destructive findings). For a case that genuinely needs the internal view, use the non-destructive findings to state exactly where a cross-section or decapsulation would be aimed and why, planning the one attempt without performing it (thermal-cameras-hotspot-detection-in-repairs).
  4. Weigh and route the step (5 minutes, desk reasoning). Decide whether the internal answer is worth the part and within safe reach, and if a destructive step is justified, state whether it would be done with proper facilities or sent to a lab, and why (root-cause-analysis-techniques).

These core steps — the classified methods, the non-destructive ladder climbed, the aimed destructive step, and the weighed decision — are tested in the Chapter Quiz at the end of this chapter, where a score of 80% is required to continue.

Key Takeaways

  • Non-destructive analysis examines a part while leaving it intact — visual, X-ray, thermal, electrical — so the part survives and the examination can be repeated, while destructive analysis consumes the part to reach what those cannot and is irreversible (x-ray-inspection-when-its-needed).
  • The governing rule is non-destructive first, always: because a destroyed part cannot be re-examined, every non-destructive method is exhausted before any destructive one, climbing a ladder from least invasive to most (failure-analysis-purpose-and-process).
  • Decapsulation exposes a chip's die and cross-sectioning reveals internal structure — the destructive rungs at the ladder's end — reached only when the non-destructive methods are exhausted and the internal answer is required (thermal-cameras-hotspot-detection-in-repairs).
  • A destructive step gets one attempt, so it is aimed by the non-destructive findings — the X-ray that located the defect telling the cut where to land — because a blind step destroys the evidence without revealing it (root-cause-analysis-techniques).
  • A destructive step is weighed against the part it spends and the hazards it carries — its aggressive acids and cutting often making it a qualified lab's task — so most repair analysis answers the question non-destructively and never destroys anything at all.

Skills Learned

After completing this section, you can:

  • Classify an analytical method as non-destructive or destructive and know which preserves the evidence.
  • Apply the non-destructive-first rule and exhaust the preserving methods before any consuming one.
  • Climb the escalation ladder from least to most invasive in the right order.
  • Use non-destructive findings to aim a destructive step rather than take it blind.
  • Judge when a destructive step is justified, and recognize the hazards that make it a lab's task.

Glossary Additions

New terms introduced in this section:

  • non-destructive analysis — the family of failure-analysis methods that examine a part while leaving it physically intact, so that the part survives the examination and the evidence is preserved. It includes visual inspection and microscopy, X-ray imaging that sees the structure inside a package without opening it, thermal imaging that locates a fault by the heat it produces, acoustic microscopy that finds internal voids and delamination, and electrical characterization that measures a part's behavior. Its defining virtue is repeatability and preservation: because the part is not consumed, an examination can be repeated, cross-checked, a second method can be applied to the same part, and the evidence remains available for later steps. This is why non-destructive analysis is always used first and exhausted before any destructive method, and why most repair-level failure analysis, whose questions the non-destructive toolkit can usually answer, never needs to destroy anything at all.
  • destructive analysis — the family of failure-analysis methods that consume or permanently alter a part in order to reveal internal features that non-destructive methods cannot reach, such as the silicon die inside a package or the internal structure of a joint or component. It includes decapsulation, which removes a chip's package to expose the die, and cross-sectioning, which mounts, cuts, grinds, and polishes a part to show a plane through its interior. Its power is access — it goes inside the part where nothing else can — and its cost is irreversibility: once a part is decapped or cross-sectioned the original is gone and cannot be re-examined, and the destructive step gets exactly one attempt. For this reason destructive analysis sits at the end of the escalation ladder, reached only when the non-destructive methods are exhausted and the internal answer is genuinely required, aimed by the non-destructive findings rather than taken blind, and, because its acids and cutting are hazardous, often performed by a qualified laboratory rather than on the bench.
  • decapsulation — a destructive analysis technique that removes the protective package of an integrated circuit to expose the silicon die inside for direct examination, done chemically by etching the plastic away with aggressive acids such as fuming nitric or sulfuric, or mechanically by grinding and prying. It is often the first destructive step taken when the non-destructive methods cannot answer a question that requires seeing the die — a cracked die, a lifted bond wire, an internal metallization defect — because it exposes the die while consuming less of the part than a full cross-section. Its chemistry is genuinely hazardous, demanding a fume hood, acid-rated protection, and proper handling and disposal, which is why for most technicians a part requiring decapsulation is sent to a qualified failure-analysis lab rather than decapped on the bench; and like every destructive step it is undertaken only after the non-destructive toolkit is exhausted and aimed by what that toolkit found.

Suggested Next Sections

Must read next:

  • Writing a Failure Analysis Report — Section 5.4 closes the chapter on communicating the result: setting down the evidence, the reasoning, the root cause, and the corrective action — including which analytical methods were used and why — in a form another technician or engineer can trust and act on.

Recommended: