Smart machining for controlling part deflection – IMaDe

This project consisted of 3 Work Packages (WP):
- WP 1: Introduction of a method coupling projection and stereo-correlation
This work was carried out as part of Théo Jovani's thesis. The thesis focused on measuring the deformations of parts during machining in a complex industrial environment. The choice was made to focus on Digital Image Correlation by applying or projecting a speckle onto the machined part.

- WP 2: Intelligent machining assembly
This work was carried out as part of Mohamed Ali Louhichi's thesis. The thesis was based on 3 axes. Axis 1 concerns the prediction of deformation using a simulation model to predict stress fields after heat treatment. Axis 2 involves the construction of a metamodel that can be used to determine the initial residual stress field by measuring the deformation of the part during machining. Axis 3 deals with the correction of simulated/measured defects. Based on the identification of the residual stress field using the methodology presented in work package 2, a correction proposal was drawn up using the so-called «intelligent« fixture, which measures the deformation of the machined part without debriding it.


- WP3: Application, design of a demonstrator
We worked on the optical measurement system used to perform image stereo-correlation based on our algorithm, and on the definition of the metamodel and the design of the intelligent assembly. Work has been carried out to couple the two instruments and obtain experimental data during the machining of the validation part. These data must now be coupled with the metamodel proposed in WP2.

The scientific work carried out as part of this project has made it possible to develop :
- A method for calculating the residual stress map from the machining of a bar layer by layer and a 2D image correlation measurement.
- A method for estimating the deformation of a plate-shaped part using speckle projection and stereo image correlation.
- A metamodel using a database of residual stress profiles deduced from heat treatment simulations. This metamodel should eventually be coupled with experimental data during machining to identify the initial residual stress map.
- An instrumented machining fixture for measuring the forces at the flanges, allowing certain deformations of the part and correction of the deformation via an additional controllable support.
This project has enabled us to carry out tests combining the measurement of 3D deformation of the part and the instrumented machining fixture. We now need to work on interpreting these measurements to identify the initial residual stress map using the metamodel. The aim is then to use these calculations to control the deformation of the part after debriding by modifying the position of the controlled support on the last machining passes.

We now need to use the data collected to identify residual stress maps using the metamodel. This work will enable us to conclude the application of our methodology to machining.
We can now envisage using this method to work on additive manufacturing, to monitor a part during printing in order to predict the residual stresses produced by thermal loading and create its digital twin.
This twin will enable us to control its behaviour during post-processing and predict its mechanical characteristics.

Jovani, T. Mesure des déformations de pièces par Corrélation d'Images Numériques pour un usinage intelligent. Matériaux. Université Clermont Auvergne, 2023. Français.

Louhichi, M. A.; Poulachon, G.; Lorong, P.; Outeiro, J.; Monteiro, E.; Cotton, D. Modeling and validation of residual stresses induced by heat treatment of AA 7075-T6 samples toward the prediction of part distortion. Machining Science and Technology. 2023, 27 (3), 247-267.

Jovani, T.; Blaysat, B.; Chanal, H.; Grédiac, M. Applying the Virtual Fields Method to measure during milling the through-thickness residual stress distribution in aluminum-alloy sheet material. Experimental Mechanics. 2023, 63 (2), 221-235.

Jovani, T.; Chanal, H.; Blaysat, B.; Grédiac, M. Direct residual stress identification during machining. Journal of Manufacturing Processes. 2022, 82, 678-688.

This project focuses on large aluminum part machining for nuclear and aerospace sectors. To meet the functional requirements, the raw parts are produced with processes which generate residual stress. The reorganization of these stresses after a removal of material induces a deformation of the part during the machining and after unclamping. This results in a non-control of the geometric quality (several millimeters) of the machined parts and the setting of recoveries additional operations.
In this project, we are interested in controlling the parts deformation during machining from an in-situ measurement of the part behavior and defining an intelligent adaptation of the clamping and the positioning of the part. This project contributes to the implementation of so-called intelligent machining adapted to the behavior of the part during machining and after unclamping.
To achieve this goal, we want to conduct research activities around two major scientific obstacles:
• The development of a quantitative imaging measurement method adapted to the machining of large parts by coupling stereo-correlation and image projection:
In this project, we want to carry out research activities to introduce a new full-field measurement method consistent with the machine tool environment and deformation kinematics. To implement this new method, formulated on a CIN approach, we will couple it with a method of projecting an image using a projector on the part surfaces. This strategy allows us to work on parts with high material removal rate. An activity of this project will focus on the choice of the projected pattern which will allow us to estimate directly by CIN the deformation of the part and displacements in the order of 0.02 mm in the machine reference frame.
The challenge then is to determine an absolute displacement of the general part shape with regard to the machining reference system and the positions of the functional surfaces in the part reference frame.
• Reduction of the dimensional defects generated by the rebalancing of the residual stresses by means of an intelligent machining clamping system:
A literature review has shown that the balancing of the machined part in the raw part, the number and the positions of the clamping system and the isostatic points of support, the values of the clamping forces as well as the machining strategy have a very significant influence on the deformation and of dimensional and geometric defects of the machined piece. Thus, we want to implement an approach consisting in acting in real time on the clamping system to constrain or deform the workpiece during machining. For this, the clamping and positioning system must be able to modify, depending on the measured quantities (clamping force, cutting forces, thickness measurement) and a pre-established strategy, the forces applied to the part and the position of its support points.
The coupling of the measured forces with a measurement of the deformation of the part and a meta-model realized within the framework of this project must make it possible to estimate in real time the final deformation of the part after unclamping. The final goal is to evolve this intelligent machining clamping system.
Even if these two scientific obstacles relate to different aspects, they are, in fact, intimately linked. The method of measuring kinematic fields and the measured stresses with the clamping system must make it possible to readjust and enrich a meta-model in order to predict the final deformation of the machined part and to change the clamping and positioning of the part.
This work will define an instrumentation and a strategy to control the quality of machined parts with residual stresses.