The inspector stops at the bolting station and points to a vehicle that rolled off the line three months ago. “Show me the documentation for bolt number seven: torque, angle profile, which tool was used, when it was last calibrated, and which worker performed the task.” The clock starts ticking the moment he says that. Either the complete chain of evidence is presented within seconds, or the search begins across separate systems—a process that raises red flags for any auditor. This article explains the twelve points that determine whether a vehicle passes or fails.
A bolting audit is not passed by whoever has the nicest reports, but by whoever can present the complete chain of evidence for every safety-critical connection within seconds. The following checklist guides you through exactly this chain, point by point.
THE MOST IMPORTANT POINTS AT A GLANCE
|
IN A NUTSHELLA bolting audit is considered successful if the complete chain of evidence for every safety-critical joint can be presented within seconds. The auditor checks three things: the correct classification of the bolting application, the completeness of the bolting data, and the unambiguous assignment of this data to the specific component. Paper records and separate systems regularly fail on the third point because the link between fastening data, tool calibration, and serial numbers is missing. The twelve items on this checklist map out precisely this chain—from the fastening application class to the required fastening data to the tool calibration record. |
A bolting audit is a standards-based inspection to verify whether safety-critical bolted joints are correctly classified, manufactured in a process-reliable manner, and fully documented. The auditor is not interested in the individual bolt, but rather in the chain of data that proves that this bolt was tightened correctly using a verified tool.
External auditors always follow the same pattern. They select a specific component, request the complete bolting history, and verify whether each required data point can be unambiguously assigned. The inspection is passed if the chain is complete, and it is at risk as soon as a single link is missing or cannot be reliably connected.
To put the order of magnitude into perspective:
| Key metric | Value |
|---|---|
| Risk classes subject to special reporting requirements | Case A and Case B |
| Retention period for production and inspection data | Over decades |
| Achievable process traceability in the event of a complaint | up to 100 percent |
| Defect costs per process step continue to | about ten times as much |
This article explores what such evidence generally looks like in an audit and explains what an audit trail in manufacturing must accomplish to pass the audit.
Auditors base their arguments on standards, not opinions. Those who are familiar with the three relevant sets of standards and clearly distinguish between their requirements can engage on equal footing during an audit.
VDI/VDE 2862. This guideline classifies bolted joints according to their risk into categories A, B, and C, and derives minimum requirements for the tightening procedure and verification documentation from these classifications. Case A covers connections whose failure could endanger life and limb and requires the highest level of safety assurance. Details on the classes are provided in the article on VDI/VDE 2862 and the bolted connection classes A, B, and C, along with their minimum requirements.
VDI/VDE 2645. This guideline governs the testing and calibration of fastening tools. An inspector will only accept a torque value if the tool that generated it has been verifiably calibrated. The calibration certificate is therefore inextricably linked to the screwdriving data record.
IATF 16949, Section 8.5.2. The automotive standard requires traceability—that is, the unambiguous assignment of process data to the individual product via the serial number. For screw connections, this means: Each screw connection data record must be assignable to a specific component.
The common thread among the three standards is the closed-loop chain consisting of correct classification, process-reliable tightening with verified tools, and unambiguous assignment. The article on VDI/VDE 2862 in practice—including its requirements and verification procedures—demonstrates how these requirements are verified in practice.
This checklist covers the entire chain of evidence that an external auditor will review during a bolting audit. Each item specifies what the auditor requires and the source of the supporting documentation. Go through the list before the audit and check each item individually.
| No. | Checkpoint | What the auditor requires | Source of Evidence |
|---|---|---|---|
| 1 | Screw Case Classification | Assignment of each connection to Case A, B, or C | Classification document in accordance with VDI/VDE 2862 |
| 2 | Tightening Procedure | Procedure appropriate for the bolting case | Process specifications and tool configuration |
| 3 | Torque Value | Actual value for each fastener within the tolerance window | Screwdriving data set with target value and tolerance |
| 4 | Angle Profile | Complete torque-angle curve | Recorded tightening curve for each joint |
| 5 | Tool Identification | Which tool created the joint | Tool ID in the fastening data record |
| 6 | Calibration Status | Valid calibration of the tool at the time | Calibration certificate in accordance with VDI/VDE 2645 |
| 7 | Timestamp | Date and time of assembly | Timestamp in the data record |
| 8 | Operator Assignment | Who performed or approved the bolting operation | Operator ID in the data record |
| 9 | Component Assignment | Unique link to the serial number | Serial number as primary key |
| 10 | Non-conforming Item (NCI) Handling | Handling of fasteners found to be non-conforming | Record of rework and approval |
| 11 | Completeness | Complete documentation of all required bolted connections | Completeness check for each order |
| 12 | Retention | Audit-proof storage for the entire retention period | Archive records with access control |
The key point: Items 3 through 9 must be able to be combined into a single, consistent data record for each bolted joint. An auditor rarely takes issue with a single value, but rather with the lack of a link between them. The article on tool management software with inspection and documentation shows exactly what the associated tool inspection and its documentation look like.
Classification is the first checkpoint and also the foundation for all subsequent ones. If a connection is classified too low, the required documentation will be missing later on, and the audit will fail due to a gap that originated during the classification process.
| Class | Risk in Case of Failure | Required Documentation |
|---|---|---|
| Case A | Danger to life and limb | Highest level of safety, complete documentation of every connection |
| Case B | Functional failure or consequential damage | Enhanced protection with documented evidence |
| Case C | Minimal impact | Basic process reliability |
The logic behind classification: The higher the risk in the event of failure, the more stringent the requirements for fastening procedures and verification. Case A requires complete documentation of every single joint, while Case C requires only basic process reliability. In practice, the common mistake is to classify components too generously into a lower category, usually to minimize effort. An inspector immediately recognizes an overly low classification because it contradicts the safety significance of the component.
An honest note is in order: Classification is an engineering decision, not a mere formality. It should be documented and justified so that, during an audit, not only the result but also the reasoning behind it is transparent.
The bolting data is at the heart of the verification. If a required field is missing or cannot be mapped, the chain is broken. The following breakdown shows which fields must be present for each bolted joint and why.
| Data Field | Purpose | Audit Relevance |
|---|---|---|
| Target Torque and Tolerance | Specification against which the check is performed | High, reference for the actual value |
| Actual torque | Value actually achieved | High, key verification of the joint |
| Torque-angle curve | Progress of the tightening process | High, demonstrates process reliability |
| Tool ID | Manufacturing tool | High, linked to calibration |
| Calibration Status | Validity of the tool inspection | High; no valid value without calibration |
| Operator ID | Person who performed or approved the inspection | Medium—proof of responsibility |
| Timestamp | Time of bolting | High, classifies the operation chronologically |
| Part serial number | Unique component assignment | High, completes the traceability chain |
The critical point is the link, not the individual field. A torque-angle curve without an associated serial number is worthless in an audit because it cannot be assigned to any specific component. Conversely, a serial number without an associated curve is incomplete documentation. The article on bolting curves, their patterns, and causes explains why the curve itself is so informative.
A common mistake that occurs time and again: Many companies record all fastening data accurately but store it separately from the serial number. During an audit, complete curves are then available but cannot be assigned to any specific component. The auditor interprets this as a lack of traceability, even though the data technically exists. The association is more important than the mere recording.
Audits rarely fail at the first stage, but usually at one of the later stages in the process. The typical gaps can be understood as a progression, ranging from the simplest to the most serious.
| Weakness | Symptom in the audit | Consequence |
|---|---|---|
| Missing Calibration | Tool without valid test certification | Torque values are not accepted |
| Missing curve | Only final value, no angular profile | Process reliability cannot be verified |
| Missing assignment | Data without serial number | Traceability interrupted |
| Incomplete data | Not all required screw connections recorded | Systematic lack of documentation |
| Missing history | Old records no longer accessible | Verification impossible in the event of a field failure |
The diagram illustrates the pattern: The lower the gap, the more fundamental the problem. A lack of calibration affects individual values, whereas a lack of historical data affects the entire ability to provide evidence over time. The most costly scenario is a lack of mapping, because it renders technically available data worthless. The article on audit readiness in manufacturing— achieved through a chain of evidence in 30 days—describes how these gaps can be systematically closed.
The difference between an audit as a major undertaking and an audit as a routine task lies in the data architecture. If fastener data, tool calibration, and serial numbers are stored in separate systems, providing proof requires a laborious manual search. If they are stored in a shared database, proof can be generated with a single query.
The closed chain of evidence is established in four steps:
The key principle: Assignment occurs at the time of creation, not during a search. This is precisely what makes the difference between a query that takes seconds and a reconstruction that takes days. The article on tool management software with inspection and documentation explains how the tool side of this process works, from inspection intervals to documentation.
CSP’s Manufacturing OS is an integrated platform on which process data, operator guidance, and audit-traceable archiving operate on a shared database. For the bolting audit, several modules of the platform work together, using the serial number as a consistent primary key.
Specifically, the Manufacturing OS maps the twelve inspection points on the checklist:
|
Because all data is based on a common foundation, audit evidence is derived not from separate systems but from a single query using the serial number. The complete fastening history of a component is thus available in seconds, including the curve, tool, and calibration status. How this integrated database makes audit readiness a byproduct of production is the core of the Audit Readiness white paper.
A bolted joint audit is a standards-based inspection to verify whether safety-critical bolted joints are correctly classified, manufactured in accordance with established processes, and fully documented. The auditor requests the complete bolting history for a specific component and verifies whether each required data point can be unambiguously assigned.
VDI/VDE 2862 classifies bolted joints according to their risk into categories A, B, and C and derives minimum requirements from this classification. VDI/VDE 2645 governs the calibration of fastening tools, and IATF 16949 Section 8.5.2 requires traceability—that is, the unambiguous assignment of data to the individual product.
Case A covers joints whose failure could result in danger to life and limb and requires the highest level of safeguards with complete documentation of each joint. Case B concerns functional failures or consequential damage and requires increased safeguards with documented evidence. Case C has a minor impact and requires only basic process reliability.
For each bolted joint, you need the target torque with tolerance, the actual value, the torque-angle curve, the tool ID, the tool’s calibration status, the operator ID, the timestamp, and the part’s serial number. It is crucial that these fields can be linked to form a consistent data set for each connection.
Not because of incorrect values, but because of a lack of association. Often, complete fastening data is available, but it is stored separately from the serial number and cannot be assigned to a specific component. The auditor interprets this as a lack of traceability, even though the data technically exists.
VDI/VDE 2645 governs the testing and calibration of fastening tools. An auditor will only accept a torque value if it can be verified that the tool used to generate it was validly calibrated at the time of fastening. The calibration record is therefore inextricably linked to the fastening data record.
In practice, production and test data for safety-critical fasteners must remain available for decades, as field failures and liability claims can occur long after the fact. Storage must be audit-traceable, ensuring that the data is tamper-proof and accessible with access controls throughout the entire retention period.
By ensuring that fastener data, tool calibration, and serial numbers are stored in a common database, with assignments made at the time of creation—not retroactively. The verification is then generated by querying the serial number rather than manually gathering information from separate systems, and the complete history is available in seconds.