Noise is not just a nuisance; it affects our health. No accepted financial data exist for the treatment costs of noise-related health problems, but several studies prove its direct relation to stress hormone levels or cardiovascular deceases, and many public organizations promote low-noise environment to protect health and comfort. Noise reduction has become major field of manufacturers’ competition: noise level data appear on domestic machines from mixers to dishwashers, brand names with ‘whisper’ or ‘silence’ are popular, aircrafts are successful due to their low noise levels. The efficient way to reduce noise is the elimination of its source that is usually vibration of some machine elements. Vibration elimination is relevant in the development of electric and self-driving cars where control panels and MEMS devices are sensitive for high-frequency excitations in the same way as high-performance machine tools are with aims at (sub)micron cutting precision. Methods of vibration reduction are based on the so-called modal testing that identifies the machines’ dynamic properties like natural frequencies. The test requires accurate broadband excitation. Commercial exciters have several drawbacks; one of these is the limited applicability for moving targets, rotating shafts. In the ERC Advanced Grant “SIREN”, a patent application was submitted and the pre-prototype of a ball shooter was constructed to excite spindles of machining centres by ball impacts. Experiments with the pre-prototype proved that the contact time is one order of magnitude shorter than that of standard impulse tests, while the force signal is near ideal: prall-free impulses with 30 kHz bandwidth were generated. Potential industrial end-users and distributors expressed interest in case a prototype is developed with accurately tuneable impact time/location together with precise online detection of impact direction for moving objects. These tasks form the work packages and deliverables of the proposal.

We propose comprehensive theoretical method development targeting a long-standing dilemma in molecular quantum simulations between controllable predictive power and affordable computational time. While the outstanding reliability of quantum chemistrys gold standard model is repeatedly corroborated against experiments, its traditional form is limited to the size of an amino acid molecule. By exploiting the short-range nature of leading interaction contributions, a handful of groups, including ours, have recently extended the reach of such quantitative energy computations up to a few hundred atoms. However, these state-of-the-art models are still too demanding and are not at all equipped to compute experimentally relevant dynamic, spectroscopic, and thermodynamic molecular properties. Thus, to break down these barriers, we will further accelerate our cutting-edge gold standard methods up to few 1000 atoms via concerted theoretical and algorithmic developments, and high-performance software design. Additionally, we will take into account biochemical, crystal, and solvent environment effects via cost-efficient embedding models. For the first time, we will also derive and implement practical approaches to compute static and dynamic observable properties for large molecules at the gold standard level. The exceptional capabilities of the new methods will enable us to study challenging chemical processes of practical importance which are not accessible with chemical accuracy for any current lower-cost alternative. We aim at modeling and understanding intricate covalent- and non-covalent interactions governing supramolecular and protein-ligand binding as well as the mechanism of organo-, organometallic, surface, and enzyme catalytic reactions. Once successful, this project we will deliver groundbreaking and open access tools for the systematically improvable and predictive quantum simulation of large molecules in realistic conditions and environments.