Simulation of adhesive flows during pressing

Schematische Darstellung - Ausbreitung von Auftragsraupen beim Verpressen

Efficient models for simulating the flow of highly viscous adhesives and sealants in manufacturing processes

1st Research objective

The first research objective is to be able to efficiently simulate the adhesive flow in joint gaps during pressing. Conventional CFD simulations often reach their limits here. The reason for this is the large aspect ratio of the application area to the joint thickness, combined with the problem of the constant need for remeshing.

By using the so-called Reynolds equation, we use an alternative approach that has been established for years in the field of lubricated friction and in which we have many years of experience [12]. In this project we transfer the mathematics of lubricated friction to adhesive flow [11] and have already shown in detail that we can simulate this flow very efficiently and realistically [8, 5]. There we can consider adhesive-specific boundary conditions such as air inclusions [7] and non-Newtonian material behaviour [9].

2nd Research objective

The second research objective is to develop suitable application patterns that ensure complete filling of the bonded joint without adhesive leakage at the edges. We have developed 3 strategies to achieve this:

  • The first strategy is based on a "reverse engineering" approach, where we reverse the time axis in the simulation. Instead of simulating the spreading of the adhesive, we analyse how the adhesive was distributed before pressing. [4]
  • In the second strategy, we track each adhesive particle and remove leaked particles from the application pattern in an iterative process. [1]
  • For the third strategy, we use artificial neural networks that are trained to determine the optimal initial distribution for a given final distribution. [2]

Finally, we experimentally demonstrated that the optimal adhesive distribution determined also optimally fills the gap in practice [3].

Optimales Auftragsmuster um ein Quadrat zu fuellen gefunden mit der Inversen Methode
Inverse method
Optimales Auftragsmuster um ein Quadrat zu fuellen gefunden mit der Iterativen Methode
Iterative method
Optimales Auftragsmuster um ein Quadrat zu fuellen gefunden mit der Neuronalen Netze Methode
Neural networks method
Experimentell gefundenes Muster welches beim Zusammenpressen ein Rechteck ergibt
Experimental method

Future research objectives

Our research objectives for the future are

  • Knowledge transfer from adhesive compression to injection
  • Simulation of the behaviour of different adhesives in a joint (keyword: biadhesives)
  • Investigation of optimal application patterns taking into account manufacturing constraints
  • Investigation of optimal application patterns with extension for hybrid joints (bonding and punching)

Selected publications

  1. Scholtes K, Flaig F, Lehne F G, Kaufmann M, Fricke H, Vallée T. Müller M. Perfect application patterns for adhesive joints found by convolutional neural networks. In progress.

  2. Kaufmann M, Flaig F, Müller M, Fricke H, Vallée T. Optimized adhesive application. International Journal of Adhesion and Adhesives 2024, p. 103620. https://doi.org/10.1016/j.ijadhadh.2024.103620

  3. Flaig, F., Fräger, T., Kaufmann, M., Vallée, T., Fricke, H., & Müller, M. A practical strategy to identify appropriate application patterns for adhesively bonded joints. Proceedings in Applied Mathematics and Mechanics 2023; e202300080. https://doi.org/10.1002/pamm.202300080

  4. Flaig F, Fräger T, Kaufmann M, Vallée T, Fricke H, Müller M. How to find the perfect application pattern for adhesively bonded joints? Journal of Advanced Joining Processes 2023;8:100147. https://doi.org/10.1016/j.jajp.2023.100147.

  5. Kaufmann M, Flaig F, Müller M, Fricke H, Vallée T. How adhesives flow during joining. International Journal of Adhesion and Adhesives 2023;122:103315. https://doi.org/10.1016/j.ijadhadh.2022.103315.

  6. Kaufmann M, Flaig F, Müller M, Fricke H, Vallée T. Do surface pretreatments for adhesives influence the squeeze flow? International Journal of Adhesion and Adhesives 2023:103362. https://doi.org/10.1016/j.ijadhadh.2023.103362.

  7. Müller M, Willenbrock S, Stahl L, Vallée T, Fricke H. Towards the efficient modelling of trapped air pockets during squeeze flow. Exp. Comput. Multiph. Flow 2023(5):29–52. https://doi.org/10.1007/s42757-021-0125-3.

  8. Kaufmann M, Flaig F, Müller M, Stahl L, Finke J, Vallée T et al. Experimental validation of a compression flow model of Non-Newtonian adhesives. The Journal of Adhesion 2022;98(14):2295–324. https://doi.org/10.1080/00218464.2021.1971081.

  9. Müller M, Finke J, Stahl L, Tong Y, Fricke H, Vallée T. Development and validation of a compression flow model of non-Newtonian adhesives. The Journal of Adhesion 2022;98(9):1260–97. https://doi.org/10.1080/00218464.2021.1895771.

  10. Müller, M; Tong, Y; Fricke, H; Vallée, T (2019): Transformation of tribological modelling of squeeze flows to simulate the flow of highly viscous adhesives and sealants in manufacturing processes. In: Proc. Appl. Math. Mech. 19 (1). https://doi.org/10.1002/pamm.201900056

  11. Müller M, Tong Y, Fricke H, Vallée T. An efficient numerical model for the evaluation of compression flow of high-viscosity adhesives. International Journal of Adhesion and Adhesives 2018;85:251–62. https://doi.org/10.1016/j.ijadhadh.2018.05.023

Selected publications of previous projects

12.   Müller, M; Ostermeyer, G-P; Bubser, F (2013): A contribution to the modeling of tribological processes under starved lubrication. In: Tribology International 64, S. 135–147. https://doi.org/10.1016/j.triboint.2013.03.011

 

Contact: Florian Flaig, M.Sc.
Deutsche Forschungsgemeinschaft
Project number: 445254897