Feedstock mixing is done to get a homogeneous mixture of polymeric material (commonly known as binder) and metal powder. Binders are used to get good moldability of metal powders. The mixture should be homogenized without any separation or segregation of materials. The blend of metal powder and binder is mixed in hot state until the metal powder particles are coated with the binder. After mixing for a specific amount of time, the mixture is taken out of the machine and allowed to cool. As it gets cold and hard, it is then ground to get feedstock pellets which are then used for molding in MIM machine. Binders are usually 15-50 vol. % of the feedstock mixture. The vol. % binder depends on the solid loading of feedstock. Eq. 1 is used to calculate the critical loading for feedstock mixing, and optimal solid loading is usually 2.5% less than critical loading.
The parameters for feedstock mixing process are mixing time, temperature, powder particles’ size & shape, formulation of binder system, shear rate, sequence of materials’ addition, and solid loading. During feedstock mixing, a rheological graph is made, which shows the change in temperature and torque with respect to mixing time. In the rheological graph shown in Figure 3, it can be seen that with every addition of input blend (binders and metal powder), the temperature of mixture decreases and the torque increases, until the solid loading point occurs where there are no significant changes in torque and temperature.
A properly mixed feedstock consists of uniformly dispersed metal powder in binder without having internal porosity or agglomerates. If it is an inhomogeneous mix, it will cause non-uniform viscosities, uneven molding and improper sintering. To get good mixing, higher shear is preferred, but not too high to damage powder particles or overheat (degrade) the binder. Small or irregular particles require more time to mix homogeneously and there are more chances of getting agglomeration. After selecting the powder size/shape, binder system and procedure of mixing, feedstock density is set to get the maximum homogeneity. The effect of input material on the torque can be seen in the Figure 3, the torque acting on mixing blades is higher when the binder isn’t in molten form and torque gets uniform, when the mix gets homogenized.
Injection molding process is carried out after feedstock is homogeneously mixed. The feedstock (at ambient temperature) is grinded first and then put in the feed hopper of MIM machine.
In the Figure 4, different parts of the MIM machine can be seen and the feedstock moves from hopper to mold in the following way.
The pellets of feedstock go to the barrel when the screw is rotating and this feedstock is melted in the barrel because of the heaters on the circumference of the barrel. As the feedstock passes through the barrel, it undergoes some frictional forces between the threads of the screw and the internal wall of the barrel. These frictional forces also raise the temperature of the barrel but its effect is relatively smaller than the temperature rise because of heaters.
After homogeneously plasticize the feedstock, it is ready for molding. The mold of MIM machine is joined to the barrel containing feedstock via sprue. When some feedstock accumulates on front of screw in barrel, it is then injected into the mold via sprue. A set amount of pressure is then applied on the feedstock in mold by the screw, which helps feedstock to properly fill the cavity of the mold. Then the screw rotates in backward direction and fills in the feedstock from hopper for next injections. In the meanwhile, the injection molded component is allowed to cool down by the help of cold water tubes in the mold. After cooling time is finished, the mold opens by the help of hydraulic system and the molded component is ejected from the cavity by ejector pins.
Debinding is done to remove the binder from the green parts which are taken out of injection molding machine. A binder usually have three components; backbone (which provides strength to the molded part), surfactant (which acts as a bridge between the binder and metal powder) and a filler part (which acts as plasticizer).
Debinding can be done in a number of ways, but the three main categories are; solvent debinding, thermal debinding and catalytic debinding. Debinding is a delicate process and debinding method depends on the filler part of binder system. The removal of binder happens in two steps; first during debinding and then during pre-sintering cycle. The filler part of binder system is removed during first stage of debinding and the rest of the binder including backbone polymer is removed during pre-sintering cycle. Backbone binder actually helps in handling the component after debinding and before sintering.
After debinding, the samples are sintered in order to get good mechanical properties. The powder particles fused together during the sintering process. Sintering temperature for a material is always less than its melting point. Figure 5 shows the progress in fusion of particles during sintering process:
After keeping the samples in sintering condition for a reasonable time, the samples are taken out with some porosity. The amount of porosity depends on the temperature and time of sintering. Porosity decreases with increase in sintering temperature and time. The engineering properties increase with increase in sintering temperature or time. But over-sintering (i.e. providing too high temperature or too long time) occurs after the properties reached the peak and starts to decline because of microstructural changes.
The surface curvature and the surface area of the powder particles provide the driving force for sintering. And during sintering, as the surface area and curvature of particles decreases, it slows the sintering processes as well. During sintering process, the shrinkage of the samples occurs. This happens because the pores created during debinding tend to close by the fusion of particles. Some pores are filled completely and some are left slightly, these left ones define the porosity of the material. During sintering, the density of the samples increases because of this volume shrinkage. The pores between the particles are filled also by the redistribution of material during sintering process.
The sample manufacturing method for this master’s thesis project was performed in the following order:
• Tapped density
• Pycnometric density
• Feedstock mixing
• Injection molding
All the stages of this methodology are briefly described.
The materials used, in this project work, were taken from Höganäs AB as well as some commercially used materials. Three different iron based powders were taken from Höganäs AB. Their names and properties are given below: 3.1.1 410
It falls in the category of ferritic-martensitic stainless steels. The general chemical composition of 410 is shown in Table 1. It is a heat treatable stainless steel. It is more hardenable than 430 & 434, shows good hardness but less corrosion resistant. It is a ferro-magnetic grade of stainless steels. 
It is a ferritic stainless steel and its general chemical composition is shown in Table 1. As it has more Cr content, so it is more corrosion resistant than 410 stainless steel. It is also ferro-magnetic and show good magnetic properties. It has good formability and is used in decorative applications. It is the most popular non-hardenable chromium stainless steel. 
It is a modified version of 430 with addition of Mo, which increases its corrosion resistance performance. The general chemical composition of 434 is shown in Table 1. It shows good oxidation resistance up to 800⁰C. It is ferro-magnetic and has good formability as well, which is because of its good ductility. Primarily used in automotive trim applications and other exterior environments. 
Reference PolyMIM 430:
The reference PolyMIM 430 has fine powder particles in it and the chemical composition is shown in Table 1, according to the feedstock datasheet.
Some properties are quite different from the 430 coarse powder taken from Höganäs AB because of the particle size difference. This material is denoted as ‘Ref 430’ in this report. 430 and 434 are ‘L’ grade stainless steels because of their Carbon content lower than 0.12%.
Stainless steel sheet material:
Stainless steel 430 sheet was cut into tensile bar samples by water jet cutting method. The dimensions of the design were taken as to equate the surface area of sintered samples. The design shown here was made with all measurements in millimeter (mm) and surface area of these samples were kept same as that of the sintered samples (which was 1280mm2). The design used for cutting the sheet metal is shown in Figure 6. The general chemical composition of sheet metal is same as that of regular 430 stainless steel and can be seen in Table 1.
410 and 434 powders were sieved by using two different sieves to get -45µm sieve cut powder. But 430 powder was already -45µm production sieved. So there was some difference in powder particle size distribution (PSD) between recently sieved (410,434) powders and already sieved 430 powder. This difference certainly created a little variation among the coarser powders and can be seen in some of the results. The particle size distribution (PSD) of coarser (-45µm) powders are shown in Figures 7, 8 & 9.
These PSD graphs show that there is some difference in median between the 430 powder (production sieved) and 410 & 434 powders (recently sieved). Although the mesh size, for the -45µm sieve used in laboratory and one used in production plant, is the same. But there may be some difference in the process parameters, which would result in this particle size variation. There might be some sort of tiny holes in laboratory sieves which might have let through the coarser particles while sieving. There is a possibility that during production sieving, some undersize powder particles might have flown out of the sieve alongside the oversize particles, causing a variation from laboratory results. These possible differences can be seen in some of the characterization results shown later in the report.
The tapped densities of the powders were measured using the “Jolting Volumeter” and the apparent densities were measured by “Hall Flowmeter”. At first apparent density was calculated by filling a density cup of known volume (i.e. 25.04cm3) without applying any pressure/force on it. Then the powder from this cup was taken out and weighed its mass. By dividing this mass by the known volume (i.e. 25.04cm3), the apparent density was calculated.
In order to measure the tapped density, ISO 3953 standard was followed. 100g of powder was weighed and put it in to the graduated cylinder. The cylinder was placed on to the holder of Jolting Volumeter. 3600 taps were set and started the machine. The machine lifts cylinder to 3mm and drop under its own weight during one tap. The machine does 250 taps in one minute . After 3600 taps the volume of the powder is measured from the graduated cylinder. By dividing 100g mass of powder by this tapped volume, tapped density is calculated. The results of apparent and tapped densities are shown in Table 2.
Gas expansion pycnometer uses helium gas (which acts as ideal gas) for the measurement of pycnometric density. It has two chambers; one used as reference and other for sample handling. The basic principle of this method is that helium under pressure takes up the volume of the chamber. The more the volume taken by helium means more the pressure of gas. So by measuring the pressure difference of the gas, the volume taken by the sample can be calculated. The mass of the sample is measured before placing it in chamber, by using mass balance. The built-in program asks for mass and does all the calculations by itself, hence giving the pycnometric density of the sample. The results obtained are shown in the Table 3. 
Some more information on gas pycnometer along with detailed results from it, are given in the Appendix I.
The results gathered from PSDs, tapped density and pycnometric density, were used for the formulation of binder system and solid loading. Then, feedstock mixing was done using a binder system and stainless steel powders. Binder system consists of filler and backbone polymer. Some amount* of binders were taken and mixed manually in a beaker. After getting a homogenous mixture, 75 %wt. of metal powder was added in the beaker. After getting it homogenized mixture in beaker, this blend was added in kneader of feedstock mixer.
Feedstock mixing was done with solid loading of 51 vol. % and 11.7 %wt. binder system. Mixing was done at 140⁰C and at 40rpm. After 15 minutes, the rest of 25 %wt. metal powder The specific binder recipe is not disclosed in this master’s thesis report because of research’s confidentiality. was added directly into the kneader. After 75 minutes of feedstock mixing, the mix was taken out and allowed to cool. After the mix cool down to ambient temperature, it was ground into pellets for injection molding process.
The injection molding parameters changes with changing binder system in feedstock mixing. So, injection molding parameters for 410, 430 and 434 were different from the parameters used for Ref430.
The coarser powder (410, 430 & 434) feedstock was injection molded with nozzle temperature at 140⁰C, cylinder temperatures of 135⁰C, 135⁰C, 125⁰C while tool temperature was kept at 25⁰C and hopper temperature was held at 25⁰C. Injection speed was kept at 5cm3/s with 16cm3 volume of feedstock for injection. Screw’s circumferential speed was kept at 140mm/s. 3000bar of max injection pressure, 1200bar of after pressure and 10bar of packing pressure was taken for these feedstocks. 20 seconds were given as cooling time before the opening of tool after the injection.
The ‘as received Ref 430 feedstock’ was injection molded with nozzle temperature at 187⁰C, cylinder temperatures of 184⁰C, 180⁰C, 175⁰C while tool temperature was kept at 50⁰C and hopper temperature was held at 35⁰C. Injection speed was kept at 20cm3/s with 15cm3 volume of feedstock for injection. Screw’s circumferential speed was kept at 166mm/s. 3000bar of max injection pressure, 600bar of after pressure and 25bar of packing pressure was taken for this feedstock. 5 seconds were given as cooling time before the opening of tool after the injection.
It should be noted that small variations in the parameters were applied accordingly, if the molding samples started to have defects after some number of injection cycles.
Solvent debinding was performed on coarse powder samples with an organic-solvent in a water bath. The water bath was kept at a temperature of 50-60⁰C and a container (Beaker) filled with organic-solvent was placed in the water bath. Solvent container was separated from the walls and bottom of water bath by a support. Water bath was used to gradually increase the temperature of solvent and keep it homogenized. The samples were placed on a wire gauge grid and each sample was separated from the other one. This wire gauge was then fully immersed into the organic-solvent container placed in water bath. The container was closed by glass lid to avoid evaporation of solvent. The samples were kept in this container for 14 hours and then taken out and dried.
For debinding Ref 430 samples, distilled water with the addition of 2% corrosion inhibitor was used as solvent. Debinding temperature was maintained 60⁰C for 10 hours in an oven. After debinding, the samples were dried for 2 hours at 100⁰C in a laboratory oven. All the results of weight-loss and dimensional changes, during debinding process, are discussed later in the Experimental results and discussion section.
Table of contents :
2 METAL INJECTION MOLDING
2.1 FEEDSTOCK MIXING
2.2 INJECTION MOLDING
3 EXPERIMENTAL METHODOLOGY
3.1.4 REFERENCE POLYMIM
3.1.5 STAINLESS STEEL SHEET MATERIAL
3.3 TAPPED DENSITY
3.4 PYCNOMETRIC DENSITY
3.5 FEEDSTOCK MIXING
3.6 INJECTION MOLDING
4 EXPERIMENTAL RESULTS AND DISCUSSION
4.1 DEBINDING AND SINTERING PROPERTIES
4.1.1 WEIGHT LOSS
4.1.2 DIMENSIONAL CHANGES
4.1.3 SINTERED DENSITY
4.2 METALLOGRAPHIC ANALYSIS
4.2.2 MICROSTRUCTURAL ANALYSIS
4.2.3 GRAIN SIZE
4.3 MECHANICAL TESTING
4.3.1 TENSILE TESTING
4.3.2 HARDNESS TESTING
4.4 CORROSION ANALYSIS
4.4.1 SALT SPRAY TEST FOR CORROSION
4.4.2 LIGHT OPTICAL MICROSCOPY (LOM)
4.4.3 SCANNING ELECTRON MICROSCOPY (SEM)
5 PROBLEMS AND DISCUSSION (POSSIBLE ERRORS)
6.1 FUTURE WORK