Connections, Anchors & Reaction Transfer Under Blast

Connections Often Govern Before the Main Member

Blast-resistant component capacity is only useful if reactions can enter the supporting structure. Welds, bolts, anchors, brackets and local support plates may experience short-duration shear, tension, prying and combined loading that differs from the quasi-static design case. Connection stiffness also controls the boundary condition of the loaded component, so connection assessment is both a strength problem and a modelling problem.

Dynamic Reaction Histories Are Not Simple Pressure Resultants

Component inertia and yielding reshape the support reaction. Peak reaction may occur before or after peak displacement, and opposite supports can carry different histories when loading is spatially non-uniform. For flexible or yielding components, the sum of instantaneous support reactions may be much lower than the applied pressure resultant because part of the load accelerates the component mass. Connection demand should therefore come from the structural response rather than a static pressure-times-area calculation.

Connection Stiffness Changes Component Capacity

A nominally fixed panel can become semi-rigid when bolt rows slip, brackets rotate or weld-adjacent material yields. This may reduce local bending moment but increase support rotation and membrane demand. Conversely, an overly rigid FE tie can suppress realistic deformation and artificially raise panel capacity. Joint stiffness should be represented at the level needed to reproduce the deformation mechanism and reaction path.

Bolted Joints Require More Than Shear Capacity

Short-duration load can produce combined bolt shear and tension, bearing, net-section demand, local plate bending and prying. Preload may delay slip but should not be assumed to eliminate it if the load exceeds friction capacity. Contact opening and re-closing can introduce impact-like force peaks. Simplified connector elements can be effective provided their axial, shear and rotational stiffness and failure limits are derived from a defensible joint model.

Anchors and Foundations Need Compatible Dynamic Assumptions

Base anchors transfer local blast reaction into foundations or supporting structure. The relevant behaviour may include anchor tension/shear interaction, base-plate bending, grout or embedment deformation and support flexibility. A perfectly fixed base can be unconservative for equipment acceleration or overly conservative for anchor force depending on the response regime. Where foundation motion is important, it should be included or bounded.

Welded Details Need Ductility and Geometry Awareness

Weld throat stress alone may not capture the governing limit under large support rotation. Weld termination, heat-affected-zone properties, local bending and constraint can control tearing. If ductile redistribution is claimed, the connection detail must actually possess rotation capacity. Brittle fracture assumptions, weld quality requirements and inspection basis should be consistent with the demanded deformation.

Reaction Transfer into Supporting Structure

The receiving structure must be capable of spreading the connection force without local crippling, punching, flange bending or fastener-row overload. Submodels should include enough surrounding material for the reaction path to develop. Applying a concentrated extracted force to an unrealistically stiff node can create nonphysical local stress and hide load distribution.

Failure Representation in Nonlinear Models

Connection failure should only be modelled explicitly when the post-failure load path matters and the failure criterion is supported by evidence. Arbitrary element deletion can shed load and energy unrealistically. For many assessments, a bounded connection capacity with separate system sensitivity cases is more defensible than an elaborate but weakly calibrated fracture model.

Verification

Check joint stiffness against hand calculations or detailed local models, verify equilibrium of transferred reactions, and inspect force-displacement histories rather than only peak force. Perform sensitivity to support stiffness, slip and preload when these could change the deformation mode. Confirm that any claimed redistribution has a physical alternate load path.

Engineering Outcome

A defensible blast connection assessment establishes both the force-transfer capability and the deformation compatibility required by the surrounding component. The report should state whether the conclusion relies on slip resistance, anchor fixity, weld ductility, prying assumptions or post-yield redistribution.

Under blast, the connection is part of the dynamic boundary condition. Its stiffness and ductility can be as important as its nominal static strength.

Combined Tension, Shear and Moment

Connection forces during blast rarely remain single-axis. Brackets and anchor groups can experience coupled shear, tension and moment as the supported component rotates. Interaction checks should reflect the connection failure mechanism rather than combine unrelated peak components that occur at different times. Time-synchronised result extraction is important for transient interaction.

Prying and Local Flexibility

Thin brackets, base plates and flanges can amplify bolt or anchor tension through prying. This flexibility can also absorb energy and reduce peak reaction. A rigid connector tied to rigid plates may therefore misrepresent both demand and deformation. Local shell/solid submodels or validated spring components can capture the relevant stiffness without modelling every thread or weld detail.

Preload, Friction and Slip Sequence

Preloaded joints may initially transfer shear by friction, then slip into bearing as demand grows. This sequence changes stiffness and can introduce impact-like re-engagement. If slip materially affects system motion, represent it with a friction/contact model or bounded stiffness states. Do not assume full friction capacity and full bearing capacity act simultaneously unless the joint mechanics support that combination.

Inspection and Quality Dependence

Connection performance can depend on weld quality, anchor installation, bolt preload and local fit-up. If the blast substantiation relies on ductile connection behaviour, those features should be controlled by inspection or manufacturing requirements. Analysis should identify the details whose as-built condition is necessary for the predicted response mechanism to remain valid.

Connection Rate Effects Should Be Mechanism Specific

Bolt material, weld metal and base material may respond differently to strain rate, and bearing or slip mechanisms may not benefit from the same dynamic strength increase as uniaxial yield. Avoid applying one global dynamic factor to the entire connection. Where rate effects are important, apply them to the constitutive behaviour actually associated with the governing limit state.

Group Effects and Load Sharing

Anchor and bolt groups may not share load uniformly when the connected plate deforms. Outer fasteners can attract tension while shear migrates after slip. A rigid spider distributing force equally can hide this behaviour. Use plate flexibility or calibrated connector stiffness when group load sharing affects margin, and check whether the most loaded fastener remains the same throughout the transient.

Reaction Filtering

Very sharp connection-force peaks may reflect local contact chatter or element-level stress waves rather than the force relevant to connection capacity. Do not smooth them away automatically. Compare raw and bandwidth-limited histories, inspect the physical duration of the peak and determine whether the connection can respond at that timescale. Report the filtering basis if filtered demand is used for acceptance.

Key takeaways

• Extract connection demand from the dynamic structural response, not pressure resultant alone.
• Represent stiffness, slip and prying when they change the deformation mechanism.
• Verify that the receiving structure and alternate load paths can actually carry redistributed reactions.