sponsored by:
27.02.2008
| TP C1 | NMR Imaging of Wavy Liquid Films |
|---|---|
| Project completed in February 2005 | |
| Subject Area: | Molecular macrochemistry, physical chemistry, NMR methods |
| Head: | Prof. Dr. Dr. h.c. Bernhard Blümich |
| Address: | Lehrstuhl für Makromolekulare Chemie
RWTH Aachen Worringer Weg 1 D-52056 Aachen |
| Phone: | +49 / 241 / 80-26420 |
| Fax: | +49 / 241 / 80-22185 |
| Email: | bluemich mc.rwth-aachen.de |
| Head: | Dr. Siegfried Stapf |
| Address: | Lehrstuhl für Makromolekulare Chemie
RWTH Aachen Worringer Weg 1 D-52056 Aachen |
| Phone: | +49 / 241 / 80-26971 |
| Fax: | +49 / 241 / 80-22185 |
| Email: | sstapfmc.rwth-aachen.de |
| Coworker: | Markus Küppers |
| Phone: | +49 / 241 / 80-26971 |
| Fax: | +49 / 241 / 80-22185 |
| Email: | mkueppersmc.rwth-aachen.de |
Project Aims
The development of two- and three-dimensional wave structures in free falling liquid filmscontains a rich phenomenology even without the additional influence of mass transfer of a
second component from the gas phase and has been empirically divided into several flow regimes.
The understanding of the strong enhancement of mass, momentum and heat transfer requires different
measurement techniques for the generation of a sufficient database which allows further
comparison with models. The direct detection of local velocities usually necessitates the
addition of tracer particles to the flowing phase. The role of NMR imaging in this project is
therefore to provide three-dimensional velocity maps with spatial resolution in a non-invasive
way. While these experiments in principle deliver six-dimensional representations of the flow
field, experimental design in conjunction with the other participating projects allows to chose
the optimal number of parameters to ensure maximum information content per experimental time.
This can be one-dimensional information in the form of averaged velocity distributions
(propagators), pure images of the film shape or reduced velocity maps of arbitrary coordinate
combinations. All measurements are performed for the whole range of laminar flow patterns up
to the crossover to the turbulent regime.
Main Achievements
A new device was constructed to give more flexibility to the researcher and to eliminate theimperfections from the device used in the past (see Fig. 1). The device has been
split into two parts, so that different distances from the start of the film to the measuring
region are adjustable. Also the speaker for an external excitation could now be placed at a
spot outside of the highest magnetic field of the NMR magnet..
Figure 1a: Photo of the new device. b: drawing of the point of start of the film.
Because of the chosen design of the point of start the film is running down the film plate in a very homogeneous way and does not form itself on the backside of the plate (see Fig. 2)
Figure 2:Image of the distribution of silicone oil in the new device. The amount of silicone is
encoded by the shades of grey (see grey scale, darker grey = higher amount).
A direct way in MRI (magnetic resonance imaging) to visualize flow is the use of tagging sequences. These sequences do monitor the deformation of an initial horizontal slice of destroyed magnetization as a function of time (see Fig. 3 & 4).
Figure 3: Selected images of a tagging sequence to visualize flow motion in a film of silicone oil M100 at Ref = 0.5 vm = 27 mm/s.
Figure 4: Selected images of a tagging sequence to visualize flow motion in a film of silicone oil M100 at Ref = 32.1 vm = 137 mm/s.
The dependence of the velocity on the direction normal to the film plate is the most relevant information to characterize the wave morphology with MRI. Theory predicts a parabolic velocity distribution for low film Reynolds numbers; the so-called Nusselt profile is indeed observed (see Fig. 5 ). For higher flow rates, waves occur spontaneously on top of a residual film; both film and waves follow the same parabolic velocity dependence and are observed as a superposition of two parabolas ( Fig. 6.). For higher Reynolds numbers, the increasing wave amplitude and their stochastic nature make it more difficult to distinguish different regimes and the residual film becomes ill-defined. Most velocity images in this research have been done with a new sequence which applies two slice selections on perpendicular cartesian axes. With this double slice selection a small portion of the film can be selectively excited. Only this small portion will give a signal during acquisition. This improves the signal to noise ratio and, when small enough, allows to perform the acquisition as two dimensional (first dimension for space encoding, second dimension for velocity encoding).
Figure 5: Velocity map of silicone oil M100 with Ref numbers 0.1; 0.2; 0.3; 0.4; and 0.5 at a distance of 38.5 cm from the start of the film. The thickness of the film is growing from 0.9 mm up to 1.5 mm. The expected Nusselt parabolas is plotted as black line into the velocity maps.
Figure 6: Velocity map of Silicon oil M100 with Ref numbers 1.1; 1,2; 1.9; 2.1; 2.4; and 3.7 at a distance of 19.5 cm from the start of the film. The thickness of the film is growing from 1.1 mm up to 2.0 mm. The expected Nusselt parabolas is plotted as black line into the velocity maps.