The best results in failure analysis are obtained when evidence can be evaluated without spoilation from repairs and poor sampling processes. The following is a basic guide for selection of samples for failure analysis and recommendations for sample removal with minimum impact on subsequent evaluation. More specific recommendations can be obtained by consulting with the MEE staff.

BEFORE BEGINNING SAMPLE REMOVAL
1. Thoroughly examine the failed components, surrounding area, and all mating components to document the failure characteristics and other notable conditions.

2. Document using 35mm photography, if possible. Photograph from several views and angles for subsequent verification of details that may not be recorded in visual inspection notes.

3. Record any operating condition data from the failure that may be available.

4. Interview operators and other witnesses for descriptions of the failure events and circumstances.

SELECTION OF SAMPLES FOR LABORATORY EVALUATION
1. If possible, remove the entire failure region with some adjacent sound material.

2. If failure sites are numerous, try to select samples of failure sites with different appearances and/or from locations with potential differences in operating conditions.

3. In many cases, it is useful to remove mating components, as well as the failed parts, e.g. collect the nuts from failed fasteners or gears and bearings with a failed shaft.

4. For large structures, obtain material sample(s) well away from the failure to allow comparison of materials properties, e.g. sample a ruptured boiler tube at the failure and near the tube sheet.

SAMPLE REMOVAL
1. Make any required cuts well away from the failure to avoid mechanical deformation and other damage that can obscure evidence at the failure site.

2. Whenever possible, make cuts with a toothed saw and no lubricant.

3. If flame cutting or abrasive cutting without lubricants must be done, make cuts far enough away from the failure site to avoid heating the material at the failure. Allow at least 3 in. for abrasive cuts and 6 in. for flame cuts, but actual distance depends on the size of the section.

4. Avoid using coolants when removing corrosion failures. Coolants will contaminate the site and obscure evidence of the corroding medium.

PACKING AND SHIPPING
1. Handle removed samples gently. Avoid touching failure sites with bare hands and do not place mating fracture surfaces together. Even light contact can smear microscopic fracture features or dislodge fragile corrosion products.

2. If samples are wet, dry in a gentle air stream before packing. In most cases, it is preferable that samples are not coated with any protective oil or other product before packing.

3. Pack samples in sealed bags or wrapped tightly in plastic. Provide cushion between samples and outside of container to avoid damage in transit.