High-fidelity modeling of interface crossing in the diffusion welding process at the polycrystalline scale
C. Godinot, E. Rigal, F. Bernard, P. Emonot, P.-E. Frayssines, L. Védie, M. Bernacki · 2025
Read High-fidelity modeling of interface crossing in the diffusion welding process at the polycrystalline scale by C. Godinot, E. Rigal, F. Bernard, P. Emonot, P.-E. Frayssines, L. Védie, M. Bernacki (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-20635.txt.
Introduction
Diffusion welding joins metals without melting them, and the joint is strongest when crystal grains grow across the original interface instead of stopping at it. Camille Godinot and six co-authors model that crossing at the scale of individual crystals, using a level-set method to simulate full two-dimensional microstructures under varied interface conditions.
They formulate two ways to measure how much crossing a simulation produces, then test and compare both against the simulated structures. The goal is control: knowing which process parameters let boundaries migrate across the weld line should help engineers secure the joint's strength and uniformity. The paper appears in 2025.
Catalog record
| CREATOR | C. Godinot, E. Rigal, F. Bernard, P. Emonot, P.-E. Frayssines, L. Védie, M. Bernacki |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-20635
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
High-fidelity modeling of interface crossing in the diffusion welding process at the polycrystalline scale C. Godinot, E. Rigal, F. Bernard, P. Emonot, P.-E. Frayssines, L. Védie, M. Bernacki Abstract Controlling the microstructure of a diffusion welded interface is a critical point to ensure optimum mechanical properties and the homogeneity of the joint. Beyond the intimate contact formation between bonded parts studied in the literature, this article focuses on the grain boundary crossing of the interface during this process and its measurement. Following this perspective, a Level-Set method has been used for full-field microstructure simulations in 2D with various interface parameters. Two crossing measurement models have been formulated, tested and discussed over the simulations. K eywords Diffusion welding, Interface crossing, Pore closure, Healing, Level-Set, High-fidelity simulation 1 Introduction Hot isostatic pressing (HIP) is widely used in industry for sintering [ 1 ] , densifying cast metals [ 2 ] , and diffusion welding (also known as diffusion bonding) [ 3 ] - a welding process in which the temperature remains below the melting point of the base materials. During this process, the parts to be welded are placed in a canister, which is then vacuumed before closure. The canister is subsequently placed in the HIP vessel, which is also closed to initiate the gas pressurization and heating. Typical HIP diffusion welding cycles last several hours, with a temperature level fixed between 70% and 80% of material melting point and a pressure range from several tens MPa to few hundred MPa. One of the primary advantages of the HIP diffusion welding process is the ability to join several metallic parts intimately with a limited impact on the microstructure. Unlike other solid-state welding processes such as explosion welding or cold welding, HIP involves isostatic stress on the parts resulting in reduced material deformation. Furthermore, in an ideal scenario, the inital interfaces become indistinguishable from the core material due to interface healing and grain boundaries migration. However, in order to maintain a high yield stress in the material e.g. for fatigue resistance, the grain size must be carefully monitored according to the Hall-Petch effect [ 4 ] , [ 5 ] . While grain boundary migration is needed for the disappearance of the interface, this mechanism will also drive a potentially unwanted excessive grain growth. This last assertion is of course specific to the case where maximizing the yield stress is the main goal. Indeed, it will not be relevant for optimizing the creep strength where a medium to high grain size is preferable [ 6 ] , for an opposite example. As illustrated in Fig. 1 , and based on experiments using Hot Uniaxial Pressing (HUP) for titanium alloys, the different stages of diffusion welding at the interface were clearly identified and described in 1954 by King and Owczarski [ 7 ] . While this process shares several features with HIP, it must be emphasized that in HUP, pressure is applied via a hydraulic press, leading to higher normal stresses on the parts and thus potentially greater strain. Nevertheless, the welding mechanisms remain the same. The first stages (see Fig. 1 ) consists in the plastic and viscoplastic deformation as the stress (either isostatic or uniaxial) is progressively imposed on the parts with the temperature rise. The initial contact points between the parts (see Fig. 1 (a)), which depend on the surface roughness, rapidly flatten, resulting in an interface with residual porosity (see Fig. 1 (b)). The hardening is limited, the metal being rapidly restored or recrystallized at the beginning of the process. As the canister has been vacuumed before HIP, the interface voids shall not be confused with trapped gas. Figure 1: a) Initial state before bonding, b) Interface formation after plastic deformation, c) Interface crossing and voids shrinkage, and d) residual voids closure - Illustrative pictures from [ 7 ] . The following stages includes both grain boundary migration by grain growth and pore closure by diffusion mechanisms (giving its name to the process). The newly formed grain boundaries move out of the interface plane, initiating crossing. At this stage, some grain boundaries may become unpinned from voids, while others moving toward voids can become pinned by these obstacles, consistent with the Smith-Zener pinning mechanism [ 8 , 9 ] , which applies to all-types of suprananometric second phase particles. In the meantime, the interface voids shrink and spheroidize, mostly by both surface and volume diffusion mechanisms at the grain boundaries (see Fig. 1 (c)). At this point, it is worth noting that while the (visco-)plastic deformation of the first stage is over once the applied stress drops below a fraction of the yield stress, grain boundary migration (and grain growth) will continue regardless of the crossing state. At the last stage, the voids intersecting a grain boundary have been closed and the remaining ones are located inside the grains. They are closed at a slower rate by volume diffusion, due to the lack of preferential diffusion paths such as grain boundaries (see Fig. 1 (d)). Although this sequential description is helpful for understanding the main mechanisms involved in diffusion welding, it is important to note that the stages often overlap. For instance, grain boundary migration through recrystallization and grain growth can begin before the initial plastic deformation is complete, and both grain boundary and volume diffusion may occur simultaneously. This step-by-step description primarily highlights the order in which the dominant mechanisms tend to emerge. Moreover precipitates can form at the interface during thermal cycle, creating new obstacles for grain boundary migration. One of the first quantitative model dedicated to diffusion bonding was proposed by Hamilton [ 10 ] who correlated the imposed temperature/stress parameters of a HUP cycle to the fraction of welded area over time for titanium. While this first analytical approach was solely based on a viscoelastic deformation model, it allowed a first quantification of the time needed to obtain a given welding ratio at the interface. Garmong et al. used the same approach [ 11 ] to develop a slice model where the local stress is calculated for each slice. These initial qualitative models were followed by numerous studies on diffusion welding, which have been progressively refined over time. Coble’s work on sintering mechanisms [ 12 ] has then served as a core foundation for the subsequent models of void closure [ 13 , 14 , 15 , 16 ] . The Derby and Wallach model [ 17 ] as well as the Hill and Wallach HUP model [ 18 ] consider seven distinct mechanisms in competition for void closure during the whole diffusion bonding process. In these approaches, a 2D cylindrical void is considered, on which each mechanism contribution is formulated depending on the initial interface geometry (no difference is made between grain boundaries connected to the void and the welding interface) and the applied stress. The ellipse major semi-axis and neck radius (itself derived from major and minor semi-axis) are computed to follow the evolution of void closure. Analytical formulations and contributions of each mechanisms for void closure are discussed in a review of Y. Takahashi and K. Inoue [ 15 ] in order to delimit the predominance domain for each one. All these state