Innovations For The Calculation Of Transmissions Up To AI

Faster, better, more user-friendly: The new version FVA Workbench 5.5 of the simulation platform for transmission systems should further accelerate development processes through insert wizards and enable even more detailed calculation results. In a short time, users can design individual gear components up to entire systems. During the calculation, insert wizards automatically suggest suitable values that can be adjusted manually. Complex planetary stages can now be modeled at lightning speed, according to the manufacturer. For different designs (e.g. Ravigneaux and Wolfrom sets), the user can model a very complex planetary gear in a short time - starting with the consideration of the installation conditions up to the automated calculation of the dependent sizes.

An overview of the new products:

Picture gallery

As a mechanical component, the gearbox is an integral part of the drive train. How to choose the right gearbox and how the integration works flawlessly is shown by our user meeting industrial gearboxes.

More information: User meeting for industrial gear units

2. New steel for gears integrated

GWJ Technology integrates steels from the company Ovako from Sweden. Ovako is a manufacturer of engineering steels for the rolling bearing and gear industry, among others. In close coordination with Ovako, GWJ implemented the two high-purity steels 159X and 159Q in its gear modules for spur gears. Both steels are based on the case-hardening steel 18CrNiMo7-6. Due to significantly lower contamination, they have modified Wöhler curves. In static terms, they are comparable to the 18CrNiMo7-6. In the fatigue strength range, however, 159Q offers around 30% higher performance for the flank load capacity and 60% higher strength for the tooth root load capacity.

3. Mathworks helps engineers build AI skills

Mathworks has released Matlab and Simulink's 2019a release. The release includes improvements in the areas of artificial intelligence (AI), signal processing and static analysis as well as new functions and bug fixes in all product families. R2019a introduces the reinforcement learning toolbox, which further improves the Matlab workflow for AI. The toolbox supports a type of machine learning with which an “agent” is trained through repeated trial-and-error interactions. “One of the biggest challenges in AI on the path from hype to production is that companies hire 'AI experts' and then try to teach them engineering skills. In contrast, Mathworks with R2019a supports engineers inexpand their AI skills - be it for the development of controls and decision-making systems with reinforcement learning, the training of deep learning models or the application of deep learning to 3D data,”said David Rich, Matlab Marketing Director at Mathworks.

4. Free software for multiscale problems

The "NG Solve" program package is now available at the Vienna University of Technology. It is up-to-date with the latest mathematical research algorithms, optimized for parallelization and can be used to solve physically complicated multiscale problems. NG Solve is provided free of charge, the team around Prof. Dr. Joachim Schöberl at the Institute for Analysis and Scientific Computing at the Vienna University of Technology, offers support to companies who want to use NG-Solve methods for complex applications and combine them with their own tools. Particular attention was paid to the parallelism of the algorithms: Thousands of processor cores are often used in large scientific computers. But this only works well if the software is perfectly designed toto break down the arithmetic into portions, which can then be solved at the same time.