The ageing and dielectric behaviour of transformer insulation is well charted; its mechanical behaviour is far less so. Yet it is the mechanical strength of pressboard, paper and laminated wood that has to survive a near short-circuit, hold clamping pressure for decades and tolerate the forces of manufacture and transport. CIGRE Technical Brochure 967 is the working group's effort to map that comparatively uncharted territory. It is a guidance document — the synthesis of a study group, not a normative standard — and it sits alongside the cellulose-ageing brochure as its mechanical counterpart: where one explains how the material ages chemically, this one explains how it carries load.
What it covers
The brochure inventories the solid insulation materials in use — kraft and thermally upgraded paper, pressboard in sheet, laminated and moulded form, laminated densified wood, aramid products and fibre-reinforced plastics — and sets out the mechanical requirements that transformer design and service impose on each. It examines how mechanical behaviour depends on direction, temperature, moisture, time and ageing, noting that fibrous and composite materials are inherently directional and that cellulose is a polymer, so its response varies with load duration. From there it identifies gaps in current standardisation and proposes new test approaches for bending, shear and elevated-temperature behaviour, then closes with a finite-element simulation framework illustrated by worked clamping-plate and conductor-bending examples.
Why it matters in practice
A recurring theme is that the test you run must match the failure mode you care about. The brochure shows, for instance, that different bending and shear geometries report systematically different strengths, that some research-grade methods are sensitive to specimen preparation, and that a simplified analytical model can over-predict stiffness where the real structure twists or delaminates. It also flags a conceptual mismatch in how thermal class is tied to temperature — to the solid alone in one framework, to the whole insulation system in another — which complicates choosing a meaningful test temperature. The explicit caution that its simulation results are illustrative and should not be over-interpreted is itself the right posture for any reader.
How we use it
This brochure informs our advisory work where mechanical condition, not fluid chemistry, is the question. When a client asks whether an aged unit can still withstand a nearby fault, it frames the mechanical-strength retention behind that judgment. When we review a third-party mechanical test report, it tells us which method suits which component geometry and which approaches remain research-grade rather than standardised. And in specification support it is useful ammunition: its tables flag where standardised elevated-temperature mechanical tests simply do not yet exist, so a client does not write a requirement that cannot be met.