MTE5881 Chap.10 Integrated Diffraction Analysis
Integrated Diffraction Analysis
The unit's research workflow combines diffraction experiments, quantitative correction, high-performance computation and technical writing around a shared materials problem. Integration means translating probe-specific observables to one structural target while retaining their corrections and uncertainty.
The literature review defines context, the progress report records intermediate analysis and the journal manuscript presents the final evidence chain. Disagreement is preserved and tested rather than silently discarded. Experimental design begins with the shared question because each method must collect observations that can eventually be compared.
XRD size trends require scale and extinction control; QCBED requires specimen screening, detector restoration, coordinates, thickness and dynamical residuals; neutron evidence retains its supported orientation or thermal role. A manuscript claim is defensible only when every reported value has a traceable experiment, correction, model translation and uncertainty.
Agreement establishes compatibility, not identity, while persistent disagreement narrows the conclusion and motivates a resolving test.
What this chapter covers
- 01
Shared research question
- 02
XRD and QCBED comparison
- 03
HPC workflow
- 04
Uncertainty propagation
- 05
Source conflict handling
- 06
Journal-manuscript claim
Worked example · free
Write an integrated evidence chain
- +2Describe the XRD scaling and zero-size extrapolation, including its uncertainty.
- +2Describe detector preparation and staged dynamical fitting behind the QCBED estimate.
- +2Translate values to the same convention and compare with uncertainty rather than raw averaging.
- +2Use neutron evidence only for the structural aspect it constrains and write a bounded claim.
Key terms
- Shared target
- A structural parameter or research question to which different observables are physically translated.
- Systematic uncertainty
- A non-random contribution from calibration, correction, modelling or experimental design.
- Bounded claim
- A conclusion limited to what all trustworthy evidence and uncertainty support.
- Evidence ledger
- A traceable table connecting each observable, correction, model translation, uncertainty and downstream claim.
- Processing state
- The declared stage of a result, such as raw, corrected, fitted or translated to a shared convention.
- Conflict ledger
- A record of competing values, their sources, operational roles and the test that could resolve disagreement.
- Resolving test
- A new calibration, observation or analysis selected because its possible outcomes distinguish candidate explanations.
- Traceability sentence
- A statement naming the experiment, corrections and model translation responsible for a reported structural value.
- Evidence lane
- A probe-specific chain preserved from raw observation through correction and uncertainty to the shared research question.
Integrated Diffraction Analysis FAQ
What does integration mean in this unit?
It means designing and interpreting different diffraction methods around one materials question, translating their observables through explicit models and comparing results with uncertainty. It does not mean erasing probe identity or averaging unlike numbers.
How should conflicting results be handled?
Preserve both values, verify common conventions, audit technique-specific controls and recompute uncertainty including systematic terms. If disagreement remains, report it and narrow the manuscript claim rather than selecting the result closest to expectation.
How do the assignments support the research workflow?
The literature review defines the problem and relevant methods, the progress report records interim practical outcomes and the journal manuscript integrates completed analysis. Their sequence mirrors context, method, result, comparison and bounded conclusion.
Exam move
Work backwards from a proposed claim. Identify the comparison that supports it, the corrected results feeding that comparison and the controls validating each result. For XRD, check scale, size dependence and extinction. For QCBED, check specimen quality, detector restoration, coordinates, thickness and dynamical-model residuals. Translate quantities to one convention before comparison and carry systematic uncertainty.
Practise writing disagreement as a result: name both readings, test probe-specific explanations and state what additional observation could resolve the conflict. Use the same method for official administrative conflicts by preserving the competing surfaces and following the live operational activity.
Draw an evidence ledger with columns for observable, processing state, correction, shared parameter, uncertainty and unresolved risk. Complete one row for every technique before writing prose. Then remove one evidence lane and test whether the proposed conclusion survives; if it collapses, state that dependence rather than strengthening the wording.
Structure a manuscript paragraph as context, method, result, comparison and bounded claim, keeping observations separate from explanations sentence by sentence. Use residual plots and size-series graphs where they expose dependencies that prose would hide. Create a conflict ledger for both measurements and course administration, recording source, date, operational role and resolving test.
Finish revision by tracing each learning outcome to a concrete decision in the workflow, from reciprocal geometry through HPC refinement to technical writing. The strongest synthesis is the smallest claim that all trustworthy evidence supports. Audit a draft by highlighting literature context, method choice, corrected observation, interpretation and limitation in different colours.
Any conclusion without all five colours needs a missing link. Build an uncertainty budget that separates random spread, calibration, correction-model dependence and translation between probe observables. When two methods disagree, assign each proposed explanation to the control capable of testing it rather than selecting the value closest to expectation.
Write captions that identify processing state, convention and uncertainty so figures remain interpretable outside the surrounding paragraph. End with a next experiment whose possible outcomes would genuinely change the claim. This makes the manuscript a traceable argument rather than a sequence of polished but disconnected results.
