Plenary Lectures

Plenary Lectures

You can download a short presentation of the Plenary Speakers: Presentation_Plenary_Speakers.pdf

Plenary Lectures abstracts

Laplace transform, regularized deconvolution and virtual thermal sensors

Denis MAILLET

University of Lorraine, France

In many cases, heat or mass transfer in a physical system can be modeled by a linear system of partial differential equations (PDE) whose coefficients do not vary with time. This is true for heat transfer in a solid, if temperature variations stay moderate, or for thermal or mass convection in a fluid, if the fluid velocity field remains stationary. If the transient excitation (heat or mass source) is unique, a Laplace transformation of the previous equations shows that the response (temperature, concentration, flux) at any point of the material system is a convolution product between this excitation and a corresponding impulse response.
Under the previous assumptions, finding this impulse response provides a reduced model which is as good as the numerical or analytical solution of the detailed PDE model. Here we show that experimental “identification” (or calibration) of this impulse response in a first step allows the use of temperature measurements, for recovering temperatures or heat fluxes at points where no sensor is present in a second “estimation” step: this corresponds to the use of a "virtual sensor". Both steps requires a deconvolution procedure, which is an ill-posed inverse problem caused by the presence of noise in the experimental signals, and a regularization is compulsory.
Applications of this two-step procedure to the characterization of heat exchangers is presented in the talk.

Covalent and Non-Covalent Interactions in Molecular Systems

Alexandre TKATCHENKO 

University of Luxembourg

This talk will concern the development of efficient, yet potentially very accurate, models to describe covalent and non-covalent (van der Waals) interactions in molecular systems. For local chemical interactions, we have developed symmetrized force-based machine learning techniques that allow to achieve the "gold standard" quantum-chemical accuracy in the description of potential-energy surfaces of mid-sized molecules [1,2]. For non-covalent interactions, we have developed coarse-grained quantum-mechanical models for interatomic potentials based on coupled harmonic oscillators [3,4]. The accuracy, efficiency, and insight that can be obtained from both approaches will be demonstrated and future directions for integrating these models into next-generation quantum force fields for complex molecular systems will be discussed. All our developments are firmly motivated by challenging experimental observations, and we make connections to experiments throughout the talk.

[1] S. Chmiela, A. Tkatchenko, H.E. Sauceda, I. Poltavsky, K.T. Schütt, and K.-R. Müller, Science Adv. 3, 1603015 (2017).
[2] S. Chmiela, H. E. Sauceda, K. R. Mueller, and A. Tkatchenko, Nature Commun. 9, 3887 (2018).
[3] J. Hermann, R. A. DiStasio Jr., and A. Tkatchenko, Chem. Rev. 117, 4714 (2017).
[4] M. Stoehr, T. Van Voorhis, and A. Tkatchenko, Chem. Soc. Rev. 48, 4118 (2019).

Non-intrusive diagnostics of micro-flows by Raman spectroscopy

Salvador MONTERO

Laboratory of Molecular Fluid Dynamics, IEM-CSIC, Serrano 121, 28006, Madrid (Spain)

emsalvador@iem.cfmac.csic.es

State-of-the-art Raman spectroscopy based on multichannel CCD detectors refrigerated by liquid N2, jointly with computer controlled spectrometers and sampling chambers, provides a powerful tool for the diagnostics of gas, liquid, and even solid micro-flows. Merits of this technique are a) its universality, in the sense that all molecular species are detectable, b) high space (~1 micron) and time resolution (~10-9 s), c) high accuracy (up to 1 %) quantitative measurements of number densities, rotational and vibrational populations and temperatures, d) wide Raynolds and Knudsen number range, e) wide spectral range (~5000 cm-1), and f) long term stability (hours). First, an overview of Raman spectroscopy methodology for flowing media will be presented. Then, a number of examples from the Laboratory of Molecular Fluid Dynamics involving gas, liquid, and solid micro-flows will be repored. These include a) supersonic jet mapping [1,2], b) structure of shock waves [3,4], c) molecular condensation [5,6], d) liquid-solid filaments [7], e) evaporative cooling [8], and f) molecular collisions [9,10]. Apart from these examples the possibilities of Raman spectroscopy for the study of confined micro-flows, transonic flow, gas-dynamic behaviour of mixtures, relaxation and transport properties in gases, as well as mapping and characterization of evaporative cooling processes will be discussed.

References

[1] G. Tejeda et al, Phys. Rev. Lett., 76, 34 (1996)
[2] B. Maté et al, J. Fluid Mech.. 426, 177 (2001).
[3] I. A. Graur et al, J. Fluid Mech., 504, 239 (2004)
[4] A Ramos et al., Phys. Rev. E, 62, 4940 (2000)
[5] G. Tejeda et al, Phys. Rev. Lett., 92, 223401 (2004)
[6] A. Ramos et al, Phys. Rev. A, 72, 053204 (2005)
[7] M. Kuehnel et al, Phys. Rev. Lett., 106, 245301 (2011)
[8] C. Goy et al, Phys. Rev. Lett., 120, 015501 (2018)
[9] J. Pérez-Ríos et al, J. Chem. Phys., 134, 174307 (2011)
[10] G. Tejeda et al, Astrophys. J. Supp. Series, 216, 3 (2015)

Development of a microcontactor for gas/liquid separation for µDMFC

Katja HAAS-SANTO 

Karlsruhe Institute of Technology (KIT) - Institute for Micro Process Engineering (IMVT)

The gas/liquid phase separation plays a key role for the many chemical processes, either to separate products or to intervene in the balance of a reaction. An example for application of orientation independent separation is in “Lab-on-a-Chip” (LOC) design-based µDirect-Methanol-Fuel-Cell (µDMFC). The overall performance is strongly dependent on the gas/liquid phase separation at the anode and cathode side, especially if the µDMFC is operated transiently with a recovery system for unused fuel (water-methanol) at the anode side.

By integrating a membrane or microsieve based micro contactor downstream into the µDMFC, the efficient removal of CO₂ from a water-methanol solution is possible. The use of membrane technology enables the separation of the two-phase flow into liquid and gas in a compact and flat device. With a membrane based micro contactor installed downstream the µDMFC, the CO₂ gas can be very efficiently be removed from the water-methanol solution.

In general, this can be achieved by using a polymeric membrane based micro contactor installed downstream of the µDMFC. However, polymeric membranes are not methanol resistant in long-term use and have a high transport resistance. In contrast, metallic or ceramic microsieves have a high thermal and chemical stability in methanol as well as a low transport resistance due to their small uniform pore diameters and length.

The gas/liquid phase separation is achieved by combination In general; the gas/liquid phase separation is achieved by using a combination of the pressure gradient as a driving force and capillary forces in the pores of the membrane acting as a transport barrier depending on the nature of it (hydrophilic/hydrophobic)

Besides process parameters as pressure gradients, temperature, and flows, the surface properties of the channels and membrane/microsieve materials play an essential role for separation performance, tailored selectivity and low system energy consumption. By a systematic study of the separation process with both polymeric membranes and metallic micro sieves the influence of several parameters have been studied. Additionally the separation efficiency (separation factor, pressure gradient, orientation and liquid loss) for different feed inlet temperatures and methanol concentration were investigated to get a better understanding of the separation process at transient working behaviour of the µDMFC.