Evidence Preservation and Documentation

Photography is the primary documentation tool in forensic engineering: it is fast, covers a wide range of scales, and produces exhibits that non-engineers can interpret in court. Forensic photography differs from ordinary photography in its purpose. The goal is an objective and complete record, which requires discipline about what to capture, how to capture it, and how the resulting files are managed.

• Use RAW format if the camera supports it. RAW preserves the sensor data without in-camera processing, giving the maximum dynamic range and allowing later adjustment of exposure without changing the underlying information. JPEG compression can obscure subtle surface features.
• Include a scale bar in every close-up photograph. Without a scale, a hairline crack on a fracture surface is indistinguishable from a millimetre-wide groove. Use a ruler or coin for informal close-ups; use a calibrated photographic scale for evidentiary photographs of fracture surfaces and small components.
• Capture a colour reference such as a Macbeth ColorChecker card, especially for rust, corrosion products, discolouration, or fire damage. Colour monitors and printers vary, and a colour reference allows accurate reproduction in reports and court exhibits.
• Never delete in-camera. Out-of-focus, over-exposed, or simply bad photographs should be kept in the original sequence. Deleted images create gaps in the file numbering that opposing counsel will highlight as potential evidence suppression.
• Maintain metadata. Camera date and time should be confirmed against an accurate reference before the investigation begins. The EXIF metadata in image files records focal length, aperture, and timestamp; these are discoverable and will be examined in significant cases.

The sequence of non-destructive examination (NDE) before destructive sampling is one of the few non-negotiable rules in forensic engineering evidence handling. NDE provides information about the evidence in its intact state. Destructive sampling provides different information (microstructural detail, compositional data, mechanical properties) that NDE cannot supply, but it changes or consumes part of the evidence irreversibly.

After NDE, if destructive sampling is needed, the investigator must decide where to sample to answer the relevant question while preserving as much of the evidence as possible for other parties. In a fracture surface examination, this typically means: photograph the fracture surface under multiple lighting conditions and angles first, then take scanning electron microscope stubs from the area of interest using a cut that does not include the fracture origin, leaving the origin area intact for the joint examination.

Spoilation of evidence is a concept that applies differently across legal systems, but the core principle is consistent: a party who destroys or materially alters evidence relevant to existing or anticipated litigation may be penalised for the resulting prejudice to the opposing party. In the United States, spoilation sanctions are governed by Federal Rule of Civil Procedure 37(e) for electronically stored information and by case law for physical evidence. Courts have awarded everything from adverse-inference instructions to dismissal of cases.

In England and Wales, the equivalent obligation arises from the Civil Procedure Rules and the courts' inherent jurisdiction. While UK courts are somewhat more reluctant to impose drastic sanctions for physical evidence loss than US federal courts, a party who destroys relevant evidence after being put on notice of potential litigation faces a real risk that the court will draw adverse inferences about what the evidence would have shown.

The forensic engineer working for any party must understand this. If you are retained by an insurer and you conduct an examination that consumes the failed component before the insured has been notified of the investigation, you have potentially created a spoilation problem for your client. Sending a spoliation letter to all interested parties before you begin examination, and waiting for acknowledgement, is the standard risk-management step.

Major forensic engineering cases routinely involve multiple claimants, insurers, contractors, manufacturers, and design professionals, each retaining their own expert. A joint inspection protocol agreed before anyone touches the evidence prevents later disputes about whether a party had a fair opportunity to examine it and who is responsible for any alteration.

1. Identify all parties with a potential interest
   Before drafting the protocol, the convening attorney (often the court-appointed neutral or the party with custody of the evidence) identifies all parties who are or may become involved in litigation related to the incident. Each potential party is notified of the inspection date and invited to attend or send an expert.
2. Agree the examination sequence
   Non-destructive examination is performed with all parties present. The sequence of examination: visual documentation first, then NDE, then sectioning and sampling by agreement. All experts observe all phases, even if only one is actively examining at any moment.
3. Divide samples equitably
   When destructive samples are taken, the protocol specifies who receives material for independent laboratory testing. The number and location of samples should be sufficient for each party's analysis without consuming more evidence than necessary.
4. Record the inspection
   A log of attendance, the sequence of actions, and the disposition of all samples is maintained by one party (agreed in advance) and distributed to all. Video recording of the inspection is increasingly common and prevents later disputes about what condition the evidence was in and what was done to it.

In England and Wales, the court has the power under CPR Part 35.7 to direct that only a single joint expert be appointed on a particular issue. Single joint experts are more common in lower-value cases and in cases where the technical issues are not genuinely disputed. In high-value cases with complex technical disputes, each party is typically allowed their own expert, and the joint inspection protocol becomes essential.

Physical engineering evidence typically spends years in storage between the incident and trial. During that time it is vulnerable to corrosion, mechanical damage, contamination, and loss. The chain of custody for engineering evidence is less formalised than in criminal forensics, but the principles are the same: document who has the evidence, where it is stored, under what conditions, and who has accessed it.

• Labelling: every item should be uniquely identified with a case number, item number, description, date of collection, and collector's name. Labels should be durable and attached in a way that does not damage the evidence.
• Packaging: fracture surfaces should be protected from contact (foam padding, tissue, separate bagging). Metal components in humid storage will corrode further; desiccant packaging or controlled storage conditions may be needed. Evidence that will be re-examined later should be packaged so it can be opened without damaging what is inside.
• Access log: a simple log recording each time an item is removed from storage, by whom, for what purpose, and when it was returned. Not as formal as a criminal exhibit log, but sufficient to rebut later claims that the evidence was accessed and altered without record.
• Photographs on return: when evidence is returned from laboratory analysis, photograph it again. Any change in condition since the last examination is recorded rather than becoming a disputed point years later at trial.

Photogrammetry and 3D laser scanning are now standard tools in major forensic engineering cases. Both methods capture the complete three-dimensional geometry of a failed component or scene with millimetre precision, creating a digital model that any expert can measure and manipulate after the fact. If the physical evidence is later lost, damaged, or modified during testing, the digital model provides the only remaining dimensional record of the evidence as found.

Structure-from-Motion (SfM) photogrammetry requires only a standard camera and readily available software (OpenDroneMap, Agisoft Metashape). The investigator takes overlapping photographs from multiple angles; the software reconstructs the three-dimensional surface from the parallax between images. The resulting point cloud and mesh can be measured in all three dimensions and sliced to produce cross-sections of the component.

Terrestrial laser scanning (TLS) provides higher accuracy and is preferred for large structures or situations where millimetre-level precision at a distance is required. FARO and Leica scanners are the standard tools; a full scene scan can be completed in minutes and delivers a point cloud accurate to 2-3 mm at distances of up to 20 metres. TLS is common in structural collapse investigations, fire-scene reconstruction, and vehicle accident reconstruction.