PSI - Issue 53

6

A. Neto et al. / Procedia Structural Integrity 53 (2024) 338–351 Alexandre de Oliveira Neto / Structural Integrity Procedia 00 (2019) 000–000

343

By choosing this failure criterion to evaluate the results of the numerical simulation, there is only one function in the used software that allows its evaluation on parts, which is MAT55 (Table 2), hence the choice. It is compatible with Shell, TShell, and Solid elements, being the latter type that was used. MAT001 (Table 3) and MAT024 (Table 4) were chosen to represent the steel impactor and aluminium supports respectively, as they were sufficient to correctly model the materials in use.

Table 3. MAT001 parameters. Non-represented parameters were left at their default value. Parameter Designation Units

Value

RO

Mass density

kg/mm 3

2,65e-03

E

Young’s modulus

[GPa]

200

PR

Poisson’s ratio

[ ]

0,3

Table 4. MAT024 parameters. Non-represented parameters were left at their default value. Parameter Designation Units

Value

RO

Mass density

kg/mm 3

2,7e-06

E

Young’s modulus

[GPa]

68,9 0,33 0,28 0,31

PR

Poisson’s ratio

[ ]

SIGY FAIL

Yield stress Failure flag

[GPa]

[ ] [ ] [ ]

C P

Strain rate parameter C Strain rate parameter P

40

5

As for the justification behind the chosen automatic general contact between interfaces (shown in Table 1), as per Dyna Support (2023), it can be said that to simulate the complex contact interactions between the impactor and the composite tube, as well as between different layers of the tube itself without predefining the contact interfaces this contact, it makes sense to use it. Not only that but also to avoid the need for mesh compatibility and conformability between the impactor and the composite tube, as well as between different parts of the tube itself. Automatic general contact can handle disjoint meshes and gaps between contact surfaces. Tobias Erhart (2011) mentions that the use of ELFORM = 1 is efficient and accurate, working even for severe formulations, but requires hourglass stabilization for an accurate transition between the elastic and plastic regime, being its value’s choice a remaining issue for the time being. Option 4 was chosen due to improving the accuracy of the simulation, especially for structural parts and also reducing the nonphysical hourglass energy that is dissipated in the system and ensuring that it is small compared to the peak internal energy for each part. 2.3. Study methodology Fundamentally, the tests were performed by incrementing the impact velocity (in numerical testing, but most importantly its relation to impact energy) until residual deformation was observed through a trial and error approach until the Tsai-Wu failure criteria limit was met. For this analysis, it made sense to have three different energy damage tiers:

 Internal damage: where the tube suffered internal invisible damage;  Intermediate level: acting as a mid-term between both tiers to show the damage evolution;  Visible external damage: where there was apparent visible damage to the outmost layer.