CBL - Campus del Baix Llobregat

Projecte llegit

Títol: Experimental study of diphasic flow during the injection stage of composite processing

Estudiants que han llegit aquest projecte:

Director/a: ZOLA, ENRICA VALERIA

Departament: ENTEL

Títol: Experimental study of diphasic flow during the injection stage of composite processing

Data inici oferta: 30-05-2019     Data finalització oferta: 30-05-2019


Estudis d'assignació del projecte:
    GR ENG SIST AEROESP
Tipus: Individual
 
Lloc de realització: ERASMUS
 
Paraules clau:
composite material, voids, diphasic flow, vacuum injection, optical fibers, conical tip, HF acid, bubbles detected.
 
Descripció del contingut i pla d'activitats:

Les bombolles porten estudiant-se fa m´es de 20 anys. En el sector aeron` autic, aquestes
bombolles apareixen durant la injecci ´o per la creaci´o d’un material compost quan l’aire es
queda atrapat. Per a que una pec¸a aeron` autica sigui certificada, ha de tenir menys d’un 2
% de cavitats.
Per tant, en aquest projecte, es fa un estudi experimental del flux dif ` asic durant la injecci ´o
en la creaci ´o d’un material compost. Aquest es duu a terme mitjanc¸ant la t `ecnica d’injecci ´o
al buit on s’han estudiat els diferents tipus de configuracions, per tal de trobar la m´es
adequada sense gaires riscos a fuga. A continuaci´ o, s’han creat fibres `optiques de punta
c`onica amb `acid HF, per la detecci´o de bombolles durant la injecci ´ o, on s’ha hagut de
tenir molta cura durant el proc´es de seguretat i creaci ´o d’aquestes. Seguidament, la fusi ´o
entre fibres ha sigut desenvolupada per tenir connexi ´o amb l’oscilloscopi i els ordinadors
que recullen la senyal que ens mostra el l `aser. Al mateix temps que el recolliment de
dades del l `aser es feia, tamb´e es gravaven instant `anies amb la c`amera, que m´es tard es
processaven amb codi Matlab per a obtenir un v´ıdeo de les imatges. Com es recollien les
dades i imatges al mateix temps, aix `o facilitava una millor la interpretaci ´o del que estava
passant en cada moment amb les bombolles. Un cop processada la informaci´o es van
dividir els tipus de senyal i es va veure quantes bombolles pod´ıem detectar en 1 minut per
veure si totes les que detectava el l `aser eren les mateixes que veiem al v´ıdeo. El programa
Matlab ens permet veure com ha sigut la detecci´o i en quin segon, des de que el l `aser ha
estat enc`es, podem veure la bombolla en el v´ıdeo.
Per altre banda, un codi Python s’ha realitzat per dibuixar les traject ` ories de les bombolles
fins arribar a la punta de la fibra. Una de les coses que es van arreglar foren les diferents
intensitats de llum que tenien cadascuna de les instant `anies que es van igualar amb la
mitja d’intensitats. A m´es a m´es, tamb´e es va fer la mitjana de 500 imatges, per tal de
tenir un fons llis on dibuixar el cam´ı de les bombolles. Finalment, algunes conclusions i
millores es van proposar per a la continuaci´o d’aquest treball en un futur.
 
Overview (resum en anglès):

The bubbles have been studied for 20 years. In the aeronautical sector, these bubbles appear
during the injection for the creation in a composite material when the air is entrapped.
For an aeronautical part certification, less than 2 % of voids is needed.
In this project, it is developed the experimental study of diphasic flow during the injection
stage of composite processing. It is applied the vacuum injection to realize it and different
set ups have been studied in order to find the more suitable one with the less leakage
probability. Then, optical fibres with a conical tip (treated with acid HF) have been created
carefully to detect the bubbles during the injection. After that, a fusion between fibres
to obtain a link with the oscilloscope and the computer was required to see the signal
of the laser. At the same time of the data recording, the camera frames were captured.
Later with a Matlab code the frames were unified to achieve a video. As the recordings
was at the same time, it was easier to interpret what was happening at every moment
with the bubbles. Once all the information was processed, the different types of signals
was divided. Then, the data signal and the video acquisition were compared to see if
the bubbles detected was the same number as in the video. The Matlab code permits
to see how the detection has been done and at which second the bubble appears in the
video, since the laser was turned on. On the other hand, a Python code has tracked the
trajectories of the bubbles until arrive to the tip.
Somethings could be improved but some of them was fixed as the different intensities in
the different frames. An average of intensities was calculated and applied in of the frames
to obtain the equal light in all the video. Furthermore, an average of 500 frames was done
in order to obtain a background without bubbles to track the trajectories. Finally, some
conclusions and improvements have been proposed to continue the study in a future.

© CBLTIC Campus del Baix Llobregat - UPC