True engineering, technical know-how is obtained when the different analyses -computer, laboratory and field based- are not treated as isolated works, but each one benefits from the others so as to achieve one single engineering process in which knowledge is neither labeled nor departmentalized. To reach this goal, the results of each one of the three phases must be compared to determine what they have in common, differences due to errors, differences due to uncertainty, etc.

It is little use to compile field, experimental and simulation data if it is not clear how to compare all the data types. Comparing data from each of these datasets, extracting information about the differences and explaining said differences is a clear strength of our hybrid engineering group.

Our knowledge spans from an understanding of how to correlate the displacement and deformation results from a static test and a simulation, to an understanding of how to correlate between simulation and operating modes, through knowledge of how to correlate real loads with those obtained in a simulation, durability of a component, etc.

One weakness of simulations is the lack of a precise understanding of certain parameters (materials, thicknesses, etc.) will mean that the real system actually behaves significantly different than the virtual one. One way to improve the precision of simulation results is to update the model. In other words, optimize the model by varying certain parameters, which are not precisely known, until the model coincides with reality.

At Aresse, we enjoy rigorous work. We trust in simulation, and put even more faith in it when it is adapted according to experimental results. This permits us to make more precise and reliable predictions.

The true potential of hybrid engineering is the possibility of working with mixed models that include real and simulation data. What logic is there in simulating a component when the information, to be obtained from the simulation, is already known? None, of course.

If this component is going to operate in conjunction with another component, and no information is available for the second component, it makes sense to work with mixed models using the real information of the known along with the virtual information of the unknown. Predictions will in turn be more accurate.

This is the Aresse method, which permits us not only to obtain better results, but also to obtain these results in less time by avoiding repetitive work.

Virtual models tend to be deterministic, which means they do not contemplate the possibility that of certain parameters, precisions, adjustments and loads may vary. The calculation returns a number, but it would be much more useful to obtain a distribution of results considering a margin of variation for the thicknesses, stresses, properties, etc. of the simulated component.

At Aresse, we specialize in precisely but realistically simulating reality. Our simulations take into consideration the possible deviation of results based on data uncertainty. We build robust designs, because they are made from models that coincide with reality, and they incorporate sensitivity analyses that consider the "what if..." scenarios.