The design of modern technology is associated with increased requirements for strength and manufacturability. Press fit connections of small size parts can withstand significant workloads, so they are widely used in mechanisms and machines. Traditionally, the calculation of press fit connections is performed according to classical formulas using a large amount of reference data. The calculations use standard recommendations that make it easier to obtain solutions with a minimum investment of time. However, the time saving limits the search to a few joints, which are not always the best in some indicators. Modeling press fit joints with different combinations of parameters leads to the formation of hundreds of potential joints. To check their performance, it is necessary to use modern computing facilities. The computer has almost unlimited possibilities for computing speed and storing arrays of information. Simulation of press fit connections on a computer allows obtaining regression metamodels, accessing the statistical significance of parameters, and determining the area of workable connections. It becomes possible to optimize the design of the joint in terms of strength, weight and assembly technology. The system for calculating and statistical analysis of press fit connections has been developed on the basis of existing standards. A program has been developed providing the joints formation and verification with an assessment of their suitability for the specified requirements.

Spur gear is used in machines to transmit rotary motion. Due to its simple design, reliability and compactness, it is widely used in technical devices — clocks, conveyors, cars. Two wheels of such a transmission have a constant gear ratio and high efficiency. The first performance criterion is the contact stress fatigue strength of the teeth. The second one is the tooth bending fatigue strength. Manual calculation of a spur gear is a difficult task due to the large number of formulas and conditions. The design area is limited to a small number of options that are not always optimal. It was proposed to expand the design boundaries to hundreds or thousands of options using a computer. Combinations of parameters (modulus, wheel width, number of teeth, material, etc.) form a variety of spur gears with different characteristics: dimensions, strength, weight, etc. The statistical analysis found hidden dependencies between characteristics and initial parameters. On the basis of regression metamodels, the relative importance of parameters for individual characteristics was established. When searching for optimal gears, the Desirability function D was applied with meta-models as arguments. On the basis of the existing standards, a system for computer-aided design of transmissions and statistical analysis of their performance has been developed. As a result of the study, the area of efficient spur gears with optimal characteristics was determined.