Advanced

Detecting Perforations in a Continuous Converted Plastic File Line

Fryking, Per and Lindqvist, Marcus (2009) In LUTEDX/TEIE EIE920 20091
Industrial Electrical Engineering and Automation
Abstract
The project was performed for FAS Converting Machinery AB (FAS) in Ystad with supervision from Industriell Elektroteknik och Automation at LTH.
FAS is a company that makes machines for producing plastic bags on a roll. The problem given was to develop a sensor for detecting the perforations in the bags to ensure that the correct amount of bags was rolled up and that the tearing was in the right place. If the machine tries to tear off where there is no perforation the fine mechanics is subjected to great stress. This was solved by detecting light penetrating the bag. The spectrum of the light is infrared due to the fact that it penetrates plastic better then visible light. To ensure that a perforation is detected and not a deformation in... (More)
The project was performed for FAS Converting Machinery AB (FAS) in Ystad with supervision from Industriell Elektroteknik och Automation at LTH.
FAS is a company that makes machines for producing plastic bags on a roll. The problem given was to develop a sensor for detecting the perforations in the bags to ensure that the correct amount of bags was rolled up and that the tearing was in the right place. If the machine tries to tear off where there is no perforation the fine mechanics is subjected to great stress. This was solved by detecting light penetrating the bag. The spectrum of the light is infrared due to the fact that it penetrates plastic better then visible light. To ensure that a perforation is detected and not a deformation in the material two things were considered. The first thing was to eliminate as many wrinkles as possible this was done by letting the plastic run over a spherical surface. The second thing was to use redundant sensors and make sure both of them triggered at the same time. What makes it difficult is that the plastic is of various density, thickness and length and a manual calibration is not applicable.
Therefore an automatic calibration algorithm had to be designed. The idea is to reduce 95% of the measured values to zero and therefore get less noise in the data. If the two sensors found a deviation at roughly the same time a signal was sent to the PLC which kept count of the produced bags.
Other problems are that the plastic might be folded up to 16 times and therefore gives a weak signal and the perforations are not perfectly overlapping each other. This was solved by designing the sensor holders like a “U” i.e. two tongues one with the IR-LED and one with the sensor. The tongue with the sensor is inserted in the fold and therefore reduces the amount of plastic to sense through. The method that was developed did manage to count the perforations in the converted plastic films. The drawback was that this method could not be implemented to fully work with all the different types of material, folding and speeds. In order to use this method, a calibration needs to be done on every single material or design change of the product produced. The automatic calibration developed to get rid of the manual calibration was not robust enough to handle all possible situations without interaction with the operator. (Less)
Please use this url to cite or link to this publication:
author
Fryking, Per and Lindqvist, Marcus
supervisor
organization
alternative title
Detektering av perforering vid kontinuerlig plastfilmstillverkning
course
EIE920 20091
year
type
H2 - Master's Degree (Two Years)
subject
publication/series
LUTEDX/TEIE
report number
5261
language
Swedish
id
1554148
alternative location
http://www.iea.lth.se/publications/MS-Theses/Full%20document/5261_full_document.pdf
date added to LUP
2014-10-07 11:16:39
date last changed
2014-10-07 11:16:39
@misc{1554148,
  abstract     = {The project was performed for FAS Converting Machinery AB (FAS) in Ystad with supervision from Industriell Elektroteknik och Automation at LTH.
FAS is a company that makes machines for producing plastic bags on a roll. The problem given was to develop a sensor for detecting the perforations in the bags to ensure that the correct amount of bags was rolled up and that the tearing was in the right place. If the machine tries to tear off where there is no perforation the fine mechanics is subjected to great stress. This was solved by detecting light penetrating the bag. The spectrum of the light is infrared due to the fact that it penetrates plastic better then visible light. To ensure that a perforation is detected and not a deformation in the material two things were considered. The first thing was to eliminate as many wrinkles as possible this was done by letting the plastic run over a spherical surface. The second thing was to use redundant sensors and make sure both of them triggered at the same time. What makes it difficult is that the plastic is of various density, thickness and length and a manual calibration is not applicable.
Therefore an automatic calibration algorithm had to be designed. The idea is to reduce 95% of the measured values to zero and therefore get less noise in the data. If the two sensors found a deviation at roughly the same time a signal was sent to the PLC which kept count of the produced bags.
Other problems are that the plastic might be folded up to 16 times and therefore gives a weak signal and the perforations are not perfectly overlapping each other. This was solved by designing the sensor holders like a “U” i.e. two tongues one with the IR-LED and one with the sensor. The tongue with the sensor is inserted in the fold and therefore reduces the amount of plastic to sense through. The method that was developed did manage to count the perforations in the converted plastic films. The drawback was that this method could not be implemented to fully work with all the different types of material, folding and speeds. In order to use this method, a calibration needs to be done on every single material or design change of the product produced. The automatic calibration developed to get rid of the manual calibration was not robust enough to handle all possible situations without interaction with the operator.},
  author       = {Fryking, Per and Lindqvist, Marcus},
  language     = {swe},
  note         = {Student Paper},
  series       = {LUTEDX/TEIE},
  title        = {Detecting Perforations in a Continuous Converted Plastic File Line},
  year         = {2009},
}