Sponsored By

MIM filling and packing dynamicsMIM filling and packing dynamics

April 1, 2002

5 Min Read
Plastics Today logo in a gray background | Plastics Today

Edited by: Michelle Maniscalco

Editor's note: At last year's PIM2TEC, a group presented the following excerpted paper that details an investigation into the filling and packing behavior of metal injection molding. The work reported here is part of a U.S. Dept. of Commerce Advanced Technology program for furthering MIM technology, which is supported by Honeywell, Ingersoll Rand, Polymer Technologies, CM Furnaces, CompAS Controls, and Penn State University. Authors of the paper include James Stevenson, Richard Roser, and Alexander Kozlov of Honeywell; and Abdessalem Derdouri and Florin Ilinca of the National Research Council of Canada.

Figure 1. In this experiment, the nozzle and mold cavity were set up with the indicated transducers to record data.



After several decades of research, the dynamics of plastics injection molding are admittedly well understood. For metal injection molding, however, a firm grasp of material behavior during the injection molding process is still developing. An experimental effort toward this goal, using an instrumented plate mold and a specific MIM material, produced results that agreed with a 3-D finite-element simulation.

First, pressure profiles measured at three locations in the mold and at two material temperature levels showed that the pressure necessary for moldfilling, as with plastics, increases sharply with increasing fill time. Second, the experiment helped to develop a procedure to determine minimum packing time.

Experimental Details

To conduct MIM molding tests, the team used a 55-ton Arburg with a 140-cu-cm shot size. The mold designed and produced for these tests featured a rectangular plate cavity 225 mm long and 25 mm wide with an adjustable thickness of 3, 4, or 5 mm. The 5-mm cavity contained removable obstacles such as a cylinder and a triangle. The mold also included a full-length knockout plate to remove parts.

As in Figure 1, the nozzle was instrumented with a Dynisco strain gauge pressure transducer and a Nanmac fast response thermocouple that extended into the melt stream. Pressure transducers located just inside the gate and at the end of the cavity were spaced 210 mm apart. The MIM compound used in the tests was PowderFlo 17-4PH U, based on gas-atomized powder with an aqueous gel binder.

Figure 2. Simulation of temperature distribution on the part surface at the instant of fill for the longest part fill time, 16.2 seconds. Cooled regions at the side and high temperatures at the end of fill are caused by fountain flow.



Filling Facts

The time required to fill the entire plate mold cavity—sprue, runner, and plate—was roughly 2.5 seconds. Once the tool was filled, a profiled pack pressure was maintained for 8 seconds, followed by 10 seconds of cooling. Also, the mold was filled in a series of runs at seven injection rates that covered the machine's limits and at two temperature setpoints—74C and 96C. Injection rates ranged from 1.6 cu cm/sec to 128 cu cm/sec.

Pressure data were collected from each of the transducers in the mold. The pressure profiles generally showed an expected minimum in injection pressure at intermediate fill times of 2 seconds (plate fill) and 3.5 seconds (cavity fill) with an injection rate of about 8 cu cm/sec.

Interestingly, this same trend has been observed for plastics as well. As fill times become shorter, the flowing material has little time to cool, so higher pressures are needed to maintain higher flow rates. At longer fill times, the increasing viscosity of the cooling material requires a higher pressure even though the flow rate is slower. This minimum pressure region is considered the optimal operating window.

Simulating MIM

The researchers simulated moldfilling for the 3-mm plate tool using a 3-D finite-element code developed by the National Research Council of Canada. For this simulation, rheological properties of the material were characterized by Datapoint Labs (Ithaca, NY).

Figure 3. Simulation of velocity distribution at the midplane for the longest part fill time. Again, areas along the side are cool, and velocity down the center of the plate is high, especially in the runner and gate areas. Figure 4. Simulation of temperature distribution on the part surface at the instant of fill for the shortest part fill time, .2 second. Shear heating causes the hot areas in the center of the plate.



The software predicted temperature distribution for the 74C setpoint and for highest part fill times (16.2 seconds), including the sprue and runner (Figure 2). For this long fill time, the material cooled (blue layer) and stopped flowing along the outside edge of the cavity. However, flow at the center of the cavity continued at a relatively higher rate (Figure 3). The team examined sintered parts molded with long fill times, and they showed flow defects consistent with this predicted flow pattern.

For the shortest fill time (.2 second), the software predicted higher surface temperatures (Figure 4). As expected, longer fill times have lower surface temperatures (28 to 76C), while the shorter fill times create higher temperatures (65 to 90C).

Minimum Pack Time

The gate-freeze method for determining pack time for plastics doesn't translate well to MIM parts. One main reason is that for MIM compounds based on aqueous gel binders, the melt becomes an elastic gel rather than a frozen solid. So the team attempted to determine minimum pack time by observing the time at which pressure inside the gate was not influenced by nozzle pressure.

To do this, the researchers held pack pressure at a 4000-psi setpoint for 2 to 64 seconds, and then dropped the setpoint to 1000 psi. They then plotted pressure at the nozzle and gate vs. time. Data indicated that after 20 seconds of pack time, nozzle pressure no longer influenced gate pressure.

This result is specific to this material and mold. However, the same type of tests could be run on other material/mold combinations to determine minimum pack time.

 

Sign up for PlasticsToday newsletter

You May Also Like