Using a Melt Flow Indexer to Test Your Pressure Transducer

The measurement of pressure is the most critical measurement in extrusion. The pressure in the extrusion die determines the output of the extruder and pressure variation causes output variation. A pressure-time trend plot enables quantitative determination of the stability of the extrusion process. Pressure feedback control enables more stable extrusion.

In addition, pressure measurement is critical to safety in extrusion. Extreme pressures can occur in the extrusion process that can lead to serious accidents. For that reason, operating an extruder without pressure measurement is extremely unsafe. Extruders should be equipped with automatic shutoff. This feature will shut off an extruder automatically when the pressure reaches a critical value.

For these reasons, it is important to make sure that your pressure transducer is working properly and capable of measuring process variation correctly. The extrusion process is susceptible to short-term pressure variation. Short-term here refers to pressure variations that occur within one second. If fact, these short-term pressure variations occur in every extrusion process. To capture these variations, the pressure transducer has to have a fast dynamic response.

This article describes a simple test method for pressure transducers that uses a melt flow indexer (MFI) modified to enable testing of pressure transducers.

The MFI is a commonly used test apparatus to determine the flow behavior of a molten plastic. It is basically a ram extruder where plastic is extruded by placing a weight on top of a plunger (see Figure 1). 

The MFI has a heated barrel. Plastic pellets are placed inside the barrel and heated until the plastic is molten and at the required test temperature. At that point, a weight is placed on top of the plunger. This weight pushes the molten plastic out of the melt index die as a single strand with a diameter of about 1-2 mm.

The MFI value is the amount of plastic extruded in 10 minutes. The amount is expressed in grams. A melt flow index of 3 indicates that 3 grams of plastic is extruded in 10 minutes. A high MFI value (for example, 10 g/10 min.) indicates a low-viscosity plastic and a low value (for example, 0.5 g/10 min.) indicates a high-viscosity plastic. The term “fractional melt” plastic is used for plastics with an MFI value less than 1.0 — these are high-viscosity plastics.

Melt Flow Indexer Modified to Test Pressure Transducers

In this study, an MFI was modified to enable testing of pressure transducers. The MFI die was removed and replaced with an adaptor with a port for a pressure transducer. The MFI was filled with a high-viscosity silicon oil with a viscosity of about 500 Pa.s. This viscosity is close to that of a typical extrusion-grade plastic. The consistency of this oil is similar to that of silly putty. 

The pressure transducer is placed in the adaptor and a weight is placed on top of the plunger. The weight determines the pressure exerted on the pressure transducer. Figure 2 shows a photo of the actual MFI discharge end. The MFI was purchased on Amazon for less than $2,000. The brand name of the MFI was not clear. The instructions suggested that the MFI was made in China.

A few issues were encountered in the testing of pressure transducers. It was found that large weights (5 kg and heavier) caused leakage because the clearance between the piston and the cylinder was too large. This was fixed by making a custom piston with a large outside diameter to prevent leakage. It was also found that the temperature sensor was not functioning in the MFI. In addition, the temperature sensor did not fit properly in the sensor well in the cylinder. As a result, the MFI could not be heated. This was fixed by installing a new RTD temperature sensor and drilling the hole of the sensor well to a diameter that ensured a good fit of the temperature sensor.

The pressure measurements were collected by an Incióna data acquisition system. Figure 3 shows a photo of the Inciona data collection system. Incióna is a specialized engineering and manufacturing services firm based in California.

The black box at the bottom is the actual data collection system. The box on top is the power supply. The measurement data is sent to the cloud. The data is collected and displayed using Grafana. This is a multiplatform, open-source analytics and interactive visualization web application. The data was downloaded as a csv file for more detailed analysis in Excel.

The pressure on the transducer is determined by the total plunger weight and the piston diameter. With the total plunger mass (Mp in kg) and the piston diameter (Dp in meter), the pressure (P in Pascal) is determined from the following expression:

P=12.5 M_p/D_P²

Five weights were available for the pressure transducer testing, 0.875 kg, 0.960 kg, 1.200 kg, 1.640 kg, and 5.000 kg. By adding a second weight on top of the first weight, the weight increased from 0.875 to 1.835, to 3.035, to 4.675, to a maximum of 9.675 kg. 

The pressure transducer that was tested was a combination pressure/temperature sensor made by Terwin, model 2076JNN10MSP268 with a range of 0-20 bar (0-3,000 psi). 

This is a “fill-free” transducer, meaning there is no fill liquid; therefore, no concern about Mercury inside the transducer. In this transducer, there is a mechanical connection between the first and second diaphragm with a cylindrical rod. This transducer was selected because of the fast response time reported by the supplier. The response time specified by the supplier was less than 2 milliseconds.

Test Procedure 1

The pressure transducer was tested by placing increasingly larger weights on the plunger. Each pressure was maintained for 1 minute. When the maximum weight (9,675 grams) was reached, the pressure was maintained for 5 minutes. Then the weight was removed one by one, again maintaining each pressure for 1 minute. Thus, the pressures were ascending in the first 300 seconds, kept constant for the next 300 seconds and descending for the last 300 seconds. 

Analysis of Test Data of Test Procedure 1

The measured pressure versus time graph is shown in Figure 4. Pressure is plotted along the vertical axis in psi, time is plotted along the horizontal axis in seconds. The pressures increase in five steps over a period of 300 seconds. Each pressure is maintained for 1 minute. The maximum pressure is reached after 300 seconds. This pressure is maintained for 300 seconds and then weights are taken off with again each pressure maintained for 1 minute.

Figure 4 shows that when the 875-gram weight is added, the pressure increases slowly over the 1-minute period and does not reach a steady value within 1 minute. A similar pattern occurs when the second 960-gram weight is added. When the third 1,200-gram weight is added, the pressure gets close to reaching a steady value after 1 minute. When the fourth 1,640-gram weight is added, the pressure almost reaches a steady value after 1 minute.

When the fifth 5,000-gram weight is added, the pressure reaches a steady value in about 30 seconds. This pressure is maintained for 300 seconds. This data shows that a steady pressure value is reached in less than 1 minute only when the 5,000-gram weight is added. For all other weights, it takes longer than 1 minute for the pressure to reach steady value.

These results indicate that for pressures below 100 psi the response time for the transducer is longer than 1 minute. The response time for pressures increasing to 200 psi the response time is reduced to about half a minute. Interestingly, the response time specified by the supplier was less than 0.002 seconds. This means that the measured response time is about 4 to 6 orders of magnitude greater than the specified values — a disturbing discrepancy!

The pressure-time curves for the descending pressure more or less follow the curves for ascending pressure with regard to the time it takes to reach steady values. In descending pressure, the steady values deviate from those obtained with ascending pressure. For instance, with a weight of 4,675 grams the steady pressure in ascending mode is about 100 psi and 130 in descending mode. That indicates considerable hysteresis in the pressure measurement.

Test Procedure 2

In the second test procedure, the pressure transducer was tested by starting with the largest weight (5,000 grams) and then placing increasingly smaller weights on the plunger. Each pressure was maintained for 1 minute. When the maximum weight (9,675 grams) was reached, the pressure was maintained for 5 minutes. Then the weights were removed one by one, again maintaining each pressure for 1 minute. The two test procedures are shown in Figure 5 as a bar chart.

Figure 6 shows pressure versus time test procedure 2, revealing a different pressure-time pattern for procedure 2 compared to procedure 1. With an increase in weight from 0 to 5,000 grams, the pressure levels off in about 20 seconds. With an increase in weight from 8,800 to 9,675 grams, the pressure levels off in about 5-10 seconds. With procedure 1, the response times ranged from 30 seconds to more than 1 minute. With procedure 2, the response times ranged from 5-20 seconds.

Clearly, the measured response times depend strongly on the sequence of weights added or removed from the plunger. It should be noted, however, that even with the lower response times measured with procedure 2, the measured values are still orders of magnitude higher than the specified values. Figure 7 shows pressure vs. time for the first 120 seconds of procedure 2.

When the 5,000-gram weight is added it takes about 2 seconds to add the weight as shown by the peak between 23 and 25 seconds. The pressure increases rapidly between 25 and 26 seconds and then levels off over the next 10 seconds. If we define the response time as the time for the sensor reading to reach 99.3% of the total step change, the response time is about 10 seconds.

The response time specified by the supplier of the pressure transducer is 0.002 seconds. This is three to four orders of magnitude faster than the measured response time using the melt flow indexer.

Future Work

The MFI tester is limited in the pressures that can be generated in the cylinder. The maximum pressure is about 200-250 psi (1.4-1.7 MPa). Higher pressures can be created using the lever principle.

Figure 8 shows how the force on the plunger can be magnified by placing a weight a distance L₁ from a pivot point with the plunger a distance L₂ from the pivot point. This arrangement enables the force to be magnified by a factor L₁/L₂. A realistic value of L₁/L₂ is 5.

The slow response time of the pressure transducer tested makes it difficult to analyze short-term pressure variations in the extrusion process. In this context, short-term variations are fluctuations that occur within one second. These short-term variations occur in every extrusion process. It is critical to have a measurement system that enables detection of these short-term variations. For that reason, it is important to obtain good data on the response time of pressure transducers used in the extrusion process. Pressure transducers with response times greater than 1 second are not suitable for determining short-term pressure variations in the extrusion process.

The melt flow indexer provides a convenient method to test pressure transducers. MFI testers are widely available and they are inexpensive. The MFI tester enables collection of meaningful pressure data in a relatively short time. It provides realistic measures of the response time of a pressure transducer. This information is critical in the interpretation of pressure data measured in an extrusion process.

ABOUT THE AUTHOR: Dr. Chris Rauwendaal is a well-known author, lecturer, researcher, entrepreneur and consultant in the field of extrusion. He holds numerous patents and has written more than 200 articles and 7 books related to extrusion, mixing, injection molding and statistical process control. A fellow of the Society of Plastics Engineers (SPE), he is the developer of the CRD, VIP and ASM mixing technologies that utilize strong elongational flow to improve mixing in extrusion and molding. Rauwendaal also developed the HHT (high-heat-transfer) extruder screw designed to improve cooling in foam tandem and other extrusion operations. In 1990, he founded and is still president of Rauwendaal Extrusion Engineering. Rauwendaal was inducted into the Plastics Hall of Fame in 2024. Contact: 530-269-1082; chris@rauwendaal.com; rauwendaal.com.