The main objective of this work is to conduct Direct Noise Computation (DNC) of high-lift wings using wall-modelled LES with LBM. For this purpose, we employ a commercial LBM software ProLB. From the qualitatively successful validation of a conventional high-lift wing (DLR/ONERA F16), in previously reported work, we carry forward the investigation with a novel Krüger flap design as an alternative high-lift device for the laminar wing concepts i.e. DLR F11. Aeroacoustically, in the absence of interacting and impinging shear layers which are present at the conventional slat, this design provides a potential for noise reduction. This can also be verified from the first LBM results, see figure 1, where the 1/3-octave band spectrum for the F16 and F11 wing is compared. However, a comparison of the Cp distribution between LBM and RANS reveals a significant under-prediction in the level of suction peaks at the Krüger high-lift wing elements. Chances are that missing turbulence from WMLES is the cause for the reduced Cp levels together with a perhaps too favourable noise reduction estimate. Inspired by the current WMLES-LBM literature [Avallone et al. 2018] on the use of e.g. distributed roughness elements to trigger the onset of turbulence, in the current work we use a RANS-informed kinematic surface roughness with the purpose of providing the missing turbulent scales and observe how the presence of roughness affects the flow. In the final presentation, we will show in addition a comparison between F16 and F11 (applied kinematic roughness) with a specific focus on the impact of roughness on the noise reduction potential.