Skip to main content

Development of a low cost in-line polymer inspection system to improve the use of recycled materials in plastics processing industry

Final Report Summary - POLYSENSE (Development of a low cost in-line polymer inspection system to improve the use of recycled materials in plastics processing industry)

Executive Summary:
The polymer industry, including producers, converters and machine manufacturers employ 1.6 million people and generate annual sales worth over €300 billion. Europe is the world leader in plastics recycling and the second worldwide plastic producer, but competence by emerging industrialised nations, such as China, is growing bigger and represents a major thread for European interests in mid and long term.

There is recently a big concern towards the recycling of wasted materials, including plastics. Despite the growth of recycled materials in the polymer sector, there are presently two main factors that prevent a higher use of recycled materials: Uncertainty in fixing final product density and quality of the final product, which is directly affected by viscosity of the blend. Yet despite this tremendous competition, many polymer processes still rely on trial and error for their optimisation, leading to the creation of enormous tonnes of scrap during adjusting periods. Current methods, which enable only indirect measurements such as temperature, screw torque and pressure, prove completely insufficient for characterising such complex behaviour.

The Polysense project has developed two new sensors to cover the needs of the extrusion industry, measuring density and viscosity in real time. The density sensor has been tested in industrial environment with satisfactory results in virgin and recycled polypropylene. The system uses the industry standard ½” port and it is able to withstand perfectly the extrusion processing conditions (temperatures up to 350 ºC and pressures up to 100 bar). The resolution of the density measurement is around 25 kg/m3 (2.5%) and the repeatability is good. The density sensor demonstrated high sensibility to the amount of fillers in the polymer and in the future could be used as a filler percentage sensor.

The viscosity sensor has only been tested in laboratory extruders, but it has demonstrated the possibility of measuring viscosity at high temperature. To manage the ultrasonic signals, a complete electronic processing system has been developed, compatible with both ultrasonic systems. Both systems are able to provide a measurement each second, despite the data processing is done in a computer. Several polymers and calibration liquids have been tested and calibrated using commercial rheometers. These measurements have been used to calibrate the ultrasonic sensors. The consortium has published in several conferences, both in polymer science and in ultrasonic technology.

Project Context and Objectives:
The main technological and scientific objective of POLYSENSE is to develop a system to increase the use of recycled material in polymer extrusion and injection processes by in-line, real-time ultrasonic measurements of density and viscosity of the blend. The determination of these variables will yield a reduction of lost material and time lost due to imperfection in the final product. The main project objectives are the following:
• To increase at least a 15% of recycled materials used in the process without affecting the product features.
• To reduce the total wasted material within the process - linked to ignoring density and viscosity values - to a 1% of the total material processed.
• To reduce the total wasted time in stopping and restarting the production process to less than 1% of the total time.

In order to achieve the above objectives, the following scientific and technological objectives and sub-objectives will be considered:
• To design and develop an ultrasonic sensor to in-line measure density and viscosity in-line in the blend in real-time.

Performance and working conditions:
- It will carry out a minimum of 1 measurement per second, and preferably 10 per second.
- It will be able to withstand temperatures up to 350 [ºC].
- It will be able to withstand pressures up to 700 [kg/cm2].
- Measurements will be obtained in PE material, including HDPE and LDPE.
- It will measure densities from 0 to 2 [g/cm3], preferably up to 3 [g/cm3]. Resolution of the measurements will be 0.01 [g/cm3], but it will be desirable to reach 0.001 [g/cm3].
- Possible viscosity values will be from 100 [Pa•s] to 100 [MPa•s], with a resolution of +/- 2%.
- Density and viscosity values will not be affected by temperature, pressure or flow variations of the polymer blend.

Project Results:
The most remarkable results coming from the Polysense project are two ultrasonic sensing systems, one to measure density and other to measure viscosity. Both systems work in a though transmission scheme (one sensor emits the signal and other one receives it), but using different properties of the ultrasonic waves.

However, the initial idea was to include both measurements in only one sensor. The ultrasonic sensors have been designed after simulate the multiple options using finite element software like ANSYS and CIVA. The ultrasonic simulations showed in a very initial stage that both the density and viscosity sensors could not be manufactured in the same sensor. Due to this limitation, the consortium decided to split the efforts in two different sensors, one for density and another one for viscosity. The most important sensor is the density sensor, since it would allow increasing the amount of recycled material used during production. Viscosity sensor will be useful to speed up the extruder tuning each time the material or the configuration of the extruder is changed.

The density sensor is able to measure the polymer density in extrusion conditions (temperatures up to 350 ºC and up to 100 bar). The density sensors were designed and manufactured using the industry standard ½” port, used commonly to place temperature and pressure sensors in extrusion systems. However, since the sensor works in a through transmission scheme, a custom adaptor to place the sensors perpendicular to the polymer flow had to be designed and manufactured. The most critical part of the sensor design was the matching layer. A matching layer is used in ultrasonic technology to increase the amount of transmitted energy from one medium to another one. In the case of polymers, the acoustic impedance is very low and to transmit acoustic energy from the sensor (in this case manufactured in stainless steel) which usually has very high impedance. Several matching layers have been tested and finally a setup using a special coating showed the best compromise between sensitivity and robustness. Polymer matching layer were also tested, but their temperature variations were too large to be considered as a possible option. The sensor was tested both in an academic environment (University of Bradford) and in an industrial environment (Multiplast). The tests in Bradford were performed using several designs, with several different matching layers. In University of Bradford different types of polymers were tested, like polyethylene, polypropylene, PVC, etc. In Multiplast, only polypropylene was used to avoid cleaning the extruder when changing the material. During all the tests performed during the project the density sensor showed the next capabilities:

• Ultrasound speed measurements very stable
• Ultrasound speed is dependant of the temperature and the screw speed (Pressure)
• Density measurement is stable, but the standard deviation is higher than in the ultrasound speed measurement
• No significant change in the density value of recycled and non recycled polypropylene
• Ultrasonic attenuation is very dependant of the amount of fillers in the polymer
• Measurements are valid only when temperature is stable

After the first measurements was clear that the pressure and the temperature were variables that will affect greatly the ultrasonic measurements, since density is also affected by temperature and pressure. To be able to compensate these issues, the adaptor used includes two extra ½” ports to measure both temperature and pressure.

One of the key points during the development was to find an easy way of calibrate the sensors. The most typical liquid used to calibrate is water, which is ok for the density range for polymers (~1000 kg/m3). However, the temperature range were the water is still liquid is quite low (<100ºC) and its viscosity is also very low. Therefore, alternative liquids were purchased to calibrate the density and viscosity sensors. These liquids, allowed calibrating the sensors at a higher range of temperature and at much higher viscosities. Moreover, University of Bradford measured the density and viscosity of all liquids at a larger range of temperatures to be able to cover all the viscosity range.

The viscosity sensor was developed during the last stage of the project. The approach of viscosity sensor is completely different to the density sensor, taking advantage of the attenuation properties of the non longitudinal waves. Like in the case of the density sensor, the first step was to simulate the possible options using ANSYS and CIVA. After this initial stage, preliminary measurements in the laboratory were performed and they showed good correlation between the attenuation and the viscosity of the material used. Again, like in the density sensor, a custom adaptor to place the viscosity sensor perpendicular to the polymer flow was manufactured, including temperature and pressure measurements. After the initial experiments, the sensor was installed in Bradford facilities to test its measuring capabilities. A Betol Davis 38 mm extruder was used to test the sensor, using different polymers with different and known viscosities. The experiments showed that the sensor response agrees with the viscosity of the material extruded. However, the some of the extruded material gets stuck in the sensor and only when the extruder is cooled down, this material goes away. This sensor has not been tested in industrial environment, but the conditions in University of Bradford are similar. The conclusions after testing the viscosity sensor were:

• The viscosity measurement is temperature dependant
• The viscosity measurement is NOT pressure dependant
• The sensitivity is low

In order to generate the high voltage pulses and to acquire the signals coming from the transducers, a complete acquisition system was designed and manufactured during the project. This electronic system was used both in the density and in the viscosity sensors. The main duties of the electronic system are:

• Generate high voltage pulses (±100V) at the central frequency of the transducers (~4MHz)
• Amplify the signal coming from the transducers adding the lowest noise possible
• Convert the analog signals into digital signals
• Capture the temperature and the pressure signals

One of the critical parts of the system is the pulse generator. It must work with high frequencies and high voltages, with very fast rise and fall times and very short activation times. In order to achieve all these requirements, the selected electronic hardware was a FPGA (Field Programmable Gate Array). The FPGAs are electronic devices, were the user configures the hardware inside the device. This devices are therefore much more fast and flexible than traditional microprocessors or DSPs. However, the development of FPGA system usually is slower and harder than with microprocessors. To avoid this situation, with Altera FPGA, you are able to generate a microprocessor inside the FPGA, a NIOS II processor. With this mixed solution, it is possible to use the flexibility of the hardware synthesizing and the commodity of a microprocessor.

Despite the FPGA system is able to acquire and process the data, the consortium agreed that processing the data in MATLAB could be much more flexible and with lower development times than processing the data in the FPGA. For this reason, a program in Labview was used to communicate a personal computer with the electronic system. This Labview software was in charge of the following duties:

• Configure the pulser parameters (number of cycles and frequency)
• Configure the analog gain for the through transmission channel
• Configure the acquisition times
• Store the signal
• Process the data

In an initial stage, the data was totally processed with MATLAB, but once a processing algorithm was developed, it was translated into MATHSCRIPT, a script language for Labview, allowing processing the data online in the Labview software. So, finally, the project achieved a totally functional sensor capable of monitoring online the density and viscosity of a polymer in a extruder.

The main problem with the injection moulding system has been the lack of space in real injection moulding machinery. Due to the high temperatures in the polymer processing systems, the ultrasonic transducers had to be separated from the heat sources (extruder or injection moulding machines) as much as possible, using buffer rods to transmit the ultrasonic waves. In the case of injection moulding machines, the available space between the ½” ports and the body of the machine is very little (a few centimetres), making impossible to maintain a temperature low enough in the ultrasonic transducers (<100 ºC).

During the project several partners have been very active publishing in polymer science or ultrasound technology events.

The exploitable foreground of the project is the following:

• The matching layer of the density sensor
• The viscosity measurement method
• The highly flexible and reconfigurable emission system
• The data processing algorithms

All the knowledge is available for the consortium SMEs. The company Rhosonics was interested in the electronic system. The SME partners decided not to apply any patent and keep the industrial secret in case in the future any of them would like to continue the research.
Potential Impact:
Europe is the second plastic producer in the world after Asia, with the 21% of world production. This sector produces 47 million tonnes per year and the production grew 1.1% in year 2011. The total turnover is arounf 89 billion € and employing 1.23 European workers. The most used polymers in the European industry are polyethylene (PE 29%), followed by polypropylene (PP 19%) and polyvinyl chloride (PVC 11%).

A very important problem of the plastic sector is the waste generated by the sector. Up to 25 million tonnes of plastic residues were generated during year 2011. But, thanks to a greater public awareness, the amount of plastic disposed in landfill decreases every year, even the fact that the production of waste grows yearly. There are four main ways of treating used plastic: Landfill, energy recovery, recovery and recycling. During 2011, recycling was the smaller treatment for plastic waste, with around 7 million tonne. Clearly, there is a need in the plastic market to increase the amount of plastic recycled, to reduce the amount of plastic landfilled.

The possibility of measuring the product density in real time while the polymer is being processed brings a new set of opportunities to polymer processors. Measuring the melt density will allow the polymer processors who use external recycled material controlling better the final quality of their products by assuring that the final product density will be within the expected values. Moreover, the Polysense density measurement sensor will provide also tools to the polymer processor to monitor special polymer blends or the fillers in the polymer, by monitoring the ultrasonic signal attenuation.

The density data of the extruded material allows the polymer processors control the quality of their products, moreover if they use recycled polymer. If they use polymers coming from recycling factories, the quality and the density of the incoming product is unknown. Usually, polymer processors mix the incoming and uncontrolled recycled polymer with virgin material. If they control the final product density, they can actuate over the gravimetric control (this is the apparatus commonly used to mix virgin and recycled polymer) and therefore adjust the amount of recycled polymer to maintain quality while decreasing costs, since recycled polymers are cheaper.

The viscosity data will help the polymer processors to setup the manufacturing process for delicate materials. There are some materials with very narrow temperature processing windows (biopolymers, some polypropylenes, some polyamides, etc), where to achieve the proper viscosity for a correct final product is very difficult. The Polysense viscosity sensor will allow the polymer processors to control online the viscosity of their product in melt state and modify the processing parameters to manufacture a better quality product. When using these materials, the amount of scrap generated by the process is very high and usually this scrap cannot be recycled because the material has been degraded. Therefore, the manufacturing costs of parts of these materials are higher, mostly due to these processing issues. The Polysense viscosity sensor would allow reducing the manufacturing costs because viscosity could be measured easily in process and therefore act in the temperature and screw rotation faster, thus reducing the scrap generated.

List of Websites: