A charge profile cannot be sold on theory alone - its cycle-life consequence must be measured, and waiting a decade for real-time results is not viable. Accelerated ageing compresses the calendar by elevating the stresses that drive degradation, but acceleration is easy to do badly: over-stress a cell and it fails by mechanisms the field will never see, yielding a pessimistic and misleading result. This paper designs credible accelerated and factorial cycle-life experiments for NiMH charge protocols, covering stress-factor selection, Arrhenius thermal acceleration, matched controls, capacity and resistance endpoints, statistical replication, and the discipline required to extrapolate field life without over-claiming - the experimental counterpart to the protocol-space analysis of Paper 24.
Acceleration should stress the same mechanisms the field profile activates (Paper 21): elevated temperature accelerates Arrhenius corrosion, increased overcharge fraction accelerates recombination-driven oxidation, and higher end-band current accelerates pressure and thermal stress. Crucially the acceleration must not introduce new failure modes - temperatures that dry electrolyte or currents that vent the cell predict nothing about a field profile that never approaches them, so stress levels stay within a justified range around the intended operating envelope.
Factorial design varies these factors together (e.g. two temperatures x three charge currents x two overcharge levels) rather than one at a time, which reveals interactions - such as temperature amplifying rate damage - and separates their contributions to degradation, directly populating the protocol-space map.
Where corrosion is the dominant mode, degradation rate approximately follows Arrhenius kinetics, doubling or tripling per 10 C over a limited range; testing at several elevated temperatures fits the activation energy and allows projection back to field temperature. Validity requires at least three temperature levels to confirm a single consistent law, plus a real-time confirmation group at field temperature that anchors the extrapolation - an accelerated result without a real-time anchor is an unverified hypothesis.
When multiple degradation modes with different activation energies coexist, an apparent Arrhenius fit can hide a change of dominant mechanism; capacity fade, resistance growth and pressure behaviour should each be examined for consistent acceleration rather than assuming one number describes all ageing.
Every protocol under test needs a matched control charged under a reference regime, with cells randomised across channels and positions to cancel channel drift and thermal gradients; cells should be drawn from the same production lot and their initial capacity/resistance matched so differences at end of test reflect the protocol, not the starting spread. Comparing a fast profile against a slow control at equal delivered energy and depth of discharge isolates the charge effect, per the fair-comparison principle of Paper 13.
Interim reference cycles - a standard IEC 61951-2 capacity and resistance check every fixed number of cycles (Paper 46) - track degradation trajectories without letting the measurement protocol itself become an ageing stress.
Define end of life before testing: typically capacity falling to 80 percent of initial, internal resistance rising past a threshold, or a pressure/vent event, whichever comes first. Tracking multiple metrics reveals the failure pathway - capacity fade with stable resistance suggests active-material loss, resistance-led fade suggests corrosion/drying (Paper 23) - and lets the experiment distinguish charge protocols that reach the same cycles-to-80 by different mechanisms and therefore different field reliability.
Post-mortem analysis (teardown, EIS per Paper 17, electrode inspection) on representative cells confirms the mechanism matches the field expectation and that acceleration did not switch failure modes.
Battery ageing has real cell-to-cell variance, so each condition needs enough replicates (commonly three to five minimum, more for reliability claims) to estimate spread and confidence bounds on cycles-to-endpoint; survival-analysis methods handle cells that have not yet failed at test end. Reporting includes all conditions, censored samples, raw trajectories and the acceleration model with its fit quality, so a life claim is auditable rather than a single best-cell number.
The first figure lays out a factorial accelerated test matrix; the second shows matched degradation trajectories and the real-time anchor used to validate extrapolation, the visual structure of a defensible life study.
A sound accelerated program converts a candidate charge profile into a predicted cycles-to-80 distribution at field conditions with stated confidence, which is exactly the evidence needed to choose between a fast and a gentle profile (Paper 24) or to justify an adaptive/optimal profile (Papers 41, 42). Weijiang supports customers with matched cells, reference cycling and historical ageing data that anchor acceleration models. The next paper steps back to compare NiMH charging against NiCd and lithium-ion as complete systems.
Weijiang Power designs and manufactures nickel-metal hydride cells, matched packs and charging-ready configurations for consumer, industrial, medical and mobility customers, and supports partners with charge-protocol guidance, IEC 61951-2 performance files, IEC 62133-1 safety evidence and charger co-validation. Share your cell format, charge rate, thermal envelope and cycle target and our engineers will specify a cell-and-charge combination that protects both runtime and service life. Review the range on the products page.
