Fatigue Assessment of Brick Masonry Arch: By Strain-Life Approach

Masonry arch bridges are integral to any national infrastructure carrying daily
passenger and freight traffic supporting the national economy, 85% of these
assets are single spans less than 10m. Improved knowledge of damage and
load-life understanding will create significant difference to safe management
and planning of interventions to asset owners.
Popular analysis methods correctly determine the four-hinge collapse mode.
All present load capacity assessments are limited to stability conditions.
Overall estimates of the collapse load do not appreciate arch behaviours
under long term cyclic loadings. Present fatigue assessment for masonry
arches relies on probabilistic models. All these models are primarily based on
tests carried out on masonry prisms. Lately researchers have started testing
masonry arches in laboratories.
Two fundamental limitations of present approaches are firstly they compare
compressive strength based on tests of undamaged specimen and secondly
numerical models ignore deformations. Recent guideline removed material
factors from test properties. Material properties do not represent present state
of a very complex multiple contributors to arch capacities. These include
cumulative damages due to repeated and increased traffic, environmental
damages, material deterioration, contribution and stiffnesses of backfill and
spandrels. Therefore, predicted load capacities are unrealistic and their
application to probabilistic models lead to unreliable outcome.
The assessment of masonry arches shall confirm that the traffic loading
does not reach levels that can cause further distress and reduce the life of
the arch (CS 454).
This research has evolved from the principles of progressive deformations, i.e., induced damages, over life of an arch. Masonry arch is tested to failure under progressive cyclic loading. Cyclic loading causes stress reversals therefore represent actual damage. Fatigue model for commonly occurring isotropic materials is well understood. Masonry is an inhomogeneous and highly anisotropic in nature. This research has simplified damage modelling by limiting to stiffness degradation. For an arch this is calibrated by comparing direct measurements to numerical results. Arch deflections are now possible to measure under passing live loads with reasonable resolution. Finally, a parametric strain-life analogy has been presented for prediction of arch life corresponding to arch capacity and resulting deflection. Additionally, strain-life approach can predict residual life of an arch by directly measuring their deformation under live loads. Damage modelling is proposed based on calibration of in-situ measurements and analytical model.
For obvious reasons the strain-life graph is based on limited laboratory tests carried out on multi-ring segmental arches. Proposal is validated through similar previous tests. A non-linear numerical modelling tools considering interface masonry modelling is recommended.
Large scale in-situ test on real scale single span arches will further improve reliability of this proposed damage-based understanding of strain-life relation. Critical understanding of deformation, damage and life will allow practicing engineers and asset owners to predict the residual life with confidence and credibly plan for future maintenance of these historic sustainable assets.