Students: Rowan Overdevest, Laurens Kok, Dave van den Berg, Jordy Rosbergen and Alex Lorier.
This production is a cooperation between BIT hotmelt Technology, Mauser Benelux BV and students of the Hague University. The students work together with BIT on a project to implement at the Mauser factory.
BIT Hotmelt Technology is a specialist in hotmelt equipment in the Benelux. They produce and service high quality applicators with customizable and innovative hotmelt products.
Mauser Benelux is a specialist in industrial packaging solutions, from composites and plastic containers to steel drums.
The project is about the production of medical waste VAT4 containers at the Mauser factory. The waste containers are sealed by hotmelt foam, a combination of glue and nitrogen. These containers need to be sealed to guarantee no smell can come out of these containers. After sealing these will be burned to dispose of the medical waste.
Fig 1. : Rim of the covers filled with hotmelt.
Problem Definition and Current situation
The glue supply is done by a robot, the robot applies the adhesive inside the rim of the cover pictured above (fig 1.). Because the supply of hotmelt and covers isn’t the same every single time the covers are susceptible to interference, meaning that the covers don’t have an even seal, which is bad for the sealing capabilities of the covers. The current setup is showed in the diagram below, as can been seen the system uses three parallel conveyor systems. A single robot is used to apply the adhesive in the three different lines. These covers are supplied by a conveyor coming from the injection moulding systems.
Fig 2. schematic of current setup.
To guarantee the seal to be perfect a checking system has to be designed to control if the seal meets the requirements imposed. To improve the rate of approval the applying of glue can be perfected.
The goals set to accomplish this project are given below:
|Checking the uniformity of the height of the foam||The uniformity is important so that every foam line is the same. The overlapping of the start and stop connection must a continuous line. Minimum 3mm-maximum 7mm||By testing|
|Checking the sealing and weight of the foam||It is important that every foam line is the same size.||By testing|
|Checking the start and stop||There must be a little overlapping in the process for a complete sealing||By testing|
|The robot has to handle in 3,33 second a cover.||The capacity of the system must be the same or faster than the current system||By calculation|
|The nozzle of the hotmelt machine must put the foam close in to the cover||There is a little lip where the nozzle must cope with, when the nozzle has an large distance to bridge, the foam can contaminate the system.||By testing|
|Nozzle cleaning by wiping||Nozzle must be clean before adding foam to the cap||By testing|
|Robot can run 3 production line at ones||There are 3 production line on each robot.||By testing|
At the tests center of BIT Hotmelt a lot of tests where done to get different parameters. Exact dates of the composition, width and height of the hotmelt foam are hard to measure. By testing the different parameters with different setting, it is clear which parameter changes. In the scheme below the different parameters are in the first column. In the second column the effect of changing the parameters is describe which parameter changes. The optimum test result is shown in the yellow cover and the blue cover is the result of Mauser. (fig1.)
Fig 3. Testing criteria
Fig 4. Test beads.
To improve the current situation to minimize the risk of shipping faulty covers to the customer a new situation has been made. In this new situation an inspection station and a separation system have been implemented. To make this proof of concept work in the robot lab at The Hague University Delft the gluing process also had to be mimicked. So the test setup will look like as sketched in figure 5.
Fig 5. Schematic of the new setup.
The inspection system has to make use of a sensor/vision system that can check the uniformity of the foam and if there are any failures in the track. This sensor also has to move along the track of the glue or the track of the glue needs to move in front of the camera. For this movement a transportation device also has to be implemented.
The separation system works together with the inspection system. When the inspection system tells the separation system to remove the next incoming cover it needs to be set to that signal. For this communication a programmable device and sensors are needed and for the removal itself also a suture and transportation device is needed.
The core of this solution is a PLC, this is a programmable logic controller, a device that controls its inputs and outputs based on a written program. This PLC is used to connect all the different components together, from input conveyor to separation station. For the test setup an old Omron PLC was used.
The gluing robot in this proof of concept is an UR10 with custom made end of arm tool that can control the hotmelt appliance supplied by BIT. This robot can deliver the speeds needed but also has the safety features needed when working with a test setup.
The inspection system used is an adaption of a XY-table, it uses the corexy mechanics found in 3D printers. This combination delivers both the accuracy needed for checking the small rim that was filled by the robot as well as the high speed needed to keep up with the robot. The checking is done by a SICK OD mini, this is a simple displacement sensor that works independently and can check to high accuracy. It is a relatively cheap way of measuring very accurately.
The separation station consists of two pistons and suctions cups, as seen below. When the horizontal piston isn’t extended the cover can be picked up from the conveyor belt. Both pistons will have to extend to drop a cover in the disapproved pile.
Fig 6. Separation station.
Implementation & Execution
The pictures below show the results of the proof of concept made by the students at the Hague University robotlab.
Fig 7. Checking system with the SICK OD sensor
Fig 8. Gluing Robot with the Hotmelt setup supplied by BIT
Fig 9. Separation station.
It was possible to get comparable results with the UR10 robot as was seen in the factory. Programming of the robot and hotmelt system gave even better results, the issue with the start and stop bead was solved. Also the weight of the adhesive inside the cover was within the limits of 12 gram maximum.
The checking system could measure the gluing bead really accurately but there where some problem connecting the arduino to the PLC. This made the checking station unreliable.
The separation system worked exactly like it should and could easily separate the good from the faulty covers.