In this article, the topic under discussion is the development of deformations in semi-rigid knee joints made of dowel type fasteners and consequences expected regarding overall deformations of timber portal frame structure. The use of semi-rigid connection resolves the problem of transportation, but development of small rotation in connection reduces the stiffness of the connection that becomes significant during service life. It is assumed that the rotational stiffness modulus Kφ (kNm∙radE−1) is the relevant characteristic of semi-rigid connection. Timber portal frame structure (span 30 m) designed with dowel type fasteners located around two circles has been subjected to different loading trials by using Dlubal software (RFEM). Results of a numerical study of portal frame model demonstrate the importance of characteristics of semi-rigid knee joints for design. It is found that developing deformations in the semi-rigid knee connection produce up to 90% bigger vertical displacement at an apex point and 96% bigger horizontal displacement comparing with the rigid knee joint model.

The use of artificial intelligence (AI) tools in university education is a phenomenon of various directions: the potential of AI tools, skills, purpose and sense of usage. Each direction is worth of working out and introducing regulatory systems and deeper investigating users’ choice and managing the process of getting, navigating and creating information by means of AI. Therefore, one of the emerging scientific challenges is students’ abilities and personalised learning experience in the use of AI. The study is focused on the usage of AI in specific courses, and namely the students of Information Technology (IT) programmes from Latvia University of Life Sciences and Technologies (LBTU) and Riga Technical University (RTU). The aim of the study is to investigate the students’ ability, need and merit to use AI in learning numerical methods, mathematics and programming. The main data collection method used is a student survey. According to the main results, it is found out that respondents when solving the programming tasks sometimes used AI, while solving mathematical tasks respondents rarely used AI. AI actually did not help to solve the mathematical tasks, while it partly helped to solve the programming tasks. The use of AI partly helped the respondents to improve the knowledge and skills of programming. Acquiring the study course Numerical Methods respondents mainly used ChatGPT, but performing practical works respondents mostly did not use AI.