*5.2. Analytical Assessment of Maximum Force Predictions (Fmax)*

A further assessment of the collected FE numerical results can be carried out with the support of two suitable analytical models of literature, namely the Eurocode 5 provisions [5] or the enhanced analytical model proposed by in [6] by Tomasi et al. Figure 12 presents the so-collected comparative data, grouped by series of TTC joints. Disregarding the joint configuration and the number/inclination of STSs, the Eurocode 5 generally manifests a weak reliability of maximum force estimates. On the other side, Figure 12 shows an improved correlation between the analytical model from [6] and the FE numerical predictions discussed herein.

**Figure 12.** Comparison of numerical (ABAQUS/Explicit) and analytical (Eurocode 5 [5] or Tomasi et al. [6]) maximum force estimates for TTC joints with inclined STSs: (**a**) S#1, (**b**) S#2, (**c**) S#3 and (**d**) S#4 joints.

As far as the percentage scatter in Equation (5) is taken into account for the analytical assessment of numerical output data, the typical result takes the form of Figure 13. For most of the examined TTC configurations, it is important to notice that the FE models in use typically manifested a marked overestimation of the analytical predictions. Following the experimental validation in Section 5.1, this suggests that both the Eurocode 5 analytical approach and the enhanced analytical model in [6] are able to only roughly capture the complex mechanical behaviour of TTC joints with inclined STSs. In any case, Figure 13 shows that the calculated percentage scatter is generally less pronounced and regular

0


20

40

60

80

100

120

S#3

FE

Analytical [5] Analytical [6]

for the FE numerical results towards the analytical formulation in [6], thus confirming the weakness of the Eurocode 5 approach.

**Figure 13.** Percentage scatter of maximum force values for TTC joints with inclined STSs (Equation (5)), as obtained from the FE numerical analyses (ABAQUS/Explicit) and by literature analytical models: (**a**) Eurocode 5 [5] and (**b**) Tomasi et al. [6].
