Abstract
Laser shock peening (LSP) is a surface treatment which improves the fatigue performance
of metallic structures by inducing compressive residual stresses. A pulsed laser vaporizes
the first layer of the component and turns the solid material into plasma. Thermal
expansion of the plasma initiates high pressure shock waves within the material. Residual
stresses are the result of local plastic deformations caused by the pressure waves. The
process is highly nonlinear and difficult to optimize based on experiments alone due to the
high number of process parameters and short time events which are hard to measure (e.g.
shock wave propagation, plasma forming). Especially, the shock wave reflection at the
back side of the specimen could influence the residual stresses and, therefore, extend the
simulated time. The material under investigation is AA2198 at different temper stages,
which shows a very pronounced rolling texture. AA2198 is an aluminum-lithium-alloy
of the third generation and finds application in light weight structures (e.g. aircraft).
Aiming to an deeper understanding of the process and an optimized residual stress
profile, a finite element model is set up. A phenomenological Johnson-Cook material
model is employed to predict the strain hardening and strain rate behavior. The work
is split into two parts; at first, experiments are used to determine the requirements for
the simulation. Experiments clarify the influence of the base layer and the influence
of the initial residual stresses depending on the thickness of the specimens as well as
the aspect of the material anisotropy. The laser is characterized by 5 J laser pulses
with a squared focus of 3mm 3mm or 1mm 1mm. Residual stresses were measured
by the incremental hole drilling method using electronic speckle pattern interferometry
(ESPI). Secondly, a finite element model is employed enabling the identification of the
laser induced pressure pulse for AA2198-T3. Laser pulse impacts were simulated with
pressure boundary conditions, hence, the influence of the phase transition of the first
material layer is neglected. The validity of the model will be shown for a LSP process
simulation of AA2198-T8. Simulation and experimental measured residual stress profiles
are for both temper stages in excellent agreement. Details about the employed numerical scheme will be presented.
