THE USE OF POWDER METALS IN DC MOTORS

George P. Gogue
Joseph J. Stupak, Jr.

G2 Consulting,
Beaverton, OR

INTRODUCTION

Motors, like other electromagnetic devices, are benefiting from a rapid development in powder metallurgy. Several grades of ferrous powders are now available in very fine particle sizes. Together with the development in die making and pressing techniques, it has become possible to produce parts whose density is very close to that of wrought iron. The process of making parts form powder metals is essentially compaction followed by sintering. There are many advantages in using powder metals, primarily in providing an economic solution to sourcing parts with complicated shapes. However, it must be understood that parts intended for the magnetic circuit can only operate at lower levels of flux-density and permeability than solid steel.

POWDER METALLURGY

Many metals and alloys are now available in powder form. Atomizing and pulverizing techniques and equipment have improved significantly, providing the user with powders of very fine sizes. Powders with particle sizes between 20 and 200 microns are available from many sources [1, 2 and 3] and, with special techniques, below 20 microns [3]. The size of the particles plays a critical role in the density and characteristics of the finished part. Various materials can be pressed into complicated shapes with extreme accuracy, and Table I gives typical tolerances of the parts as sintered (an essential primary operation) and as sized (an optional secondary operation) [6].

Even though various materials can be used for structural and other purposes, ferrous powders are of special interest to designers and will be our primary focus.

THE PROCESS

The first step is selecting the appropriate material on the basis of its intended use. The following parameters of the finished part should serve as a guide when discussing the material with the powder supplier:

1. Minimum acceptable tensile strength.

2. Minimum value of saturation flux-density.

3. Maximum value of core losses at particular values of flux-density and frequency.

While the above parameters are not determined by the choice of material alone and are greatly influenced by the process also, the material is, nevertheless, a critical first step.

The various metals in their recommended percentages are mixed thoroughly and additives and lubricant may be added for process improvement. This is done, preferably, at the facilities of the powder supplier.

The next step is compacting and it can be done at either supplier or user facilities. The most common method is mechanical compaction by applying a significant amount of pressure (30 to 60 ton/sq-in) on the powder in a suitably constructed die.

Compacting can be done at room temperature or in heated dies and molds. Compaction is followed by sintering in which the part is heated to the temperature specified by the powder supplier, causing the adjacent surfaces of particles to bond together. Sintering is done in a total or partial Nitrogen environment or even in a vacuum. For steel, the temperature is typically between 2000°F and 2500°F, and may be maintained at those levels up to 2 hours.

Many other operations can be performed on the part after sintering but they are all optional and tailored for the specific application. Examples of such operations are sizing, heat treating, machining, coating, etc. Sizing is a pressing operation to establish dimensional tolerances, allowed by the porous nature of powder metal parts. Oil impregnation and coatings of various kinds can also be applied to the finished parts after sintering.

BASIC DESIGN RULES

To achieve the full benefit of powder metal technology in the construction of dc motors, a few simple rules must be followed in the design process [4]:

1. Select for this technology only the parts which have forms too complicated for other machining operations. It is not economical to make simple parts from powder metals.

2. The shape of the part must allow the construction of strong molds and dies to withstand high pressures for a large number of cycles.

3. Avoid sharp corners and thin walls because they are difficult to protect from breakage.

4. Design the part so that it is possible to eject it from the mold after pressing. This means leaving undercuts, grooves and some holds for machining operations after sintering.

It is important to stress that the high cost of tooling can only be offset by the prospect of large volume production to result in low part cost. The initial cost of tooling can be as much as five times what is needed for punched or machined parts.

MECHANICAL PROPERTIES

Table II lists some mechanical parameters of selected materials. The tensile strength of the finished parts is of critical importance, and heat treatment is seen from the data to improve this parameter significantly [4, 7].

Percentage elongation is a measure of how brittle the finished parts are. Materials with percentages less than 2% are considered very brittle and would result in high breakage if subjected to secondary operations after compaction. Heat treatment appears to reduce this value while increasing the tensile strength.

Many other alloys are available in powder form: some containing trace amounts of certain metals to produce specific characteristics. These requirements can be discussed with the powder suppliers while discussing other properties of the powder. For example, increasing amounts of carbon (up to 0.8%) improves the strength of the finished part. Alternatively, controlled amounts of phosphorus and silicon can substantially improve the magnetic characteristics, as will be seen in the next section.

MAGNETIC PROPERTIES

As designers of electromagnetic devices and in particular, motors, our main interest lies in the magnetic behavior of the parts made from powder metals. It has always been of concern that such parts do not have the equivalent density of steel available in the form of bars or laminations. While this is still true, high levels of compaction can result in a high percentage of the theoretical density. Notwithstanding this disadvantage, powder metals can have a variety of additives included in their mix, e.g., iron phosphate which significantly improves the high-frequency properties [1].

In selecting the appropriate material, the parameters of concern are the saturation flux-density, resistivity, coercive force, remanence, and core losses. It has been reported [9] that the saturation flux-density (measured at 25 Oersted) and resistivity are determined solely by the alloy used. The coercive force (and hence the permeability) and remanence, however, are determined by the process followed in preparing the particular parts. The processes of mechanical compacting and sintering both influence the size and shape of the space between particles and hence the magnetic properties. For example, the higher the temperature is during sintering, the higher the saturation flux-density achievable. In the case of Ancorsteel 1000B [1], the flux-density is reported to increase from about 4 kG to 14 kG when the sintering temperature is raised from 2000°F to 2500°F, [9]. The coercive force appears to drop from approximately 3 to 1 Oersted as a result of these changes in temperature conditions. The coercive force is also reduced by an increased proportion of hydrogen to nitrogen in the environment during sintering. This is probably due to the removal of some of the oxides of the metals in the powder, forming water vapor [9]. Figure 1 shows the B-H curve for a typical powder metal part [1].

Small amounts of phosphorus result in higher values of saturation flux-density (at 25 Oe) and improvement is more significant at low sintering temperatures. The coercive force is virtually half the amount measured for F-0008 (Table II) for the same temperature conditions [9]. The resistivity of the material, however, is adversely influenced by the addition of phosphorus. This value is typically about 20 micro Ohm cm versus about 50 for silicon steel. It therefore, slows down the rate of decay of the magnetic field in the material when the external field is removed. In addition to a higher remanence, the core losses are also higher as a consequence. Figure 2 shows the variation in resistivity with phosphorus content.

Alternatively, small amounts of phosphorus are used in the form of a layer of iron phosphate on the iron particles to control the core losses. This amount ranges between 0.001%. and 0.2%, and is applied in a proprietary process [1]. This addition is particularly advantageous for high frequency applications where the core losses increase very rapidly.

Silicon steel is commonly used in the design and construction of electric motors. The most usual form is thin laminations. In preparing the powder alloy, 2-6% silicon can be added. The additional amount of silicon can cause dimensional changes in the range of 2-3%. at the high temperatures during sintering [9]. The resulting deformation must then be corrected by machining after sintering. Another disadvantage resulting from increasing amounts of silicon is a corresponding reduction in the saturation flux-density. On the other hand, the core losses are reduced with increasing percentages of silicon.

Table IIIa gives values of some magnetic parameters for commonly available powder metals [8]. Table IIIb lists the same parameters for wrought materials for comparison.

The core losses experienced with each material are a function of the usual parameters, i.e., flux-density and frequency of field reversal. Additionally, in the case of powder metal parts, the losses are a function of the density of the material, the coating on the particles (if any), the sintering process and possibly the heat treatment that may follow. Because of the large number of variables, it is difficult to get consistent data as in the case of laminated steel. Table IV gives the core loss/cubic cm for one powder metal material [2] and two other materials in lamination form. The values were obtained at 10kG.

Figure 3 shows the core losses at 5 kG for three materials, silicon steel laminations assembly, 1008 mild steel assembly and typical powder metal compacted (but not sintered) into a similar toroid [1].

Depending on the operating frequency, the core losses with powder metal parts can be higher or lower than those of silicon steel laminations. However, they appear to be consistently lower than the losses experienced with mild steel.

Figures 4a and 4b show the core losses of powder metal parts which are sintered after compaction. In (a) the losses are plotted against induction (at 50 Hz) showing a rapid increase at high values of flux-density. In (b) the losses are plotted against frequency (at 6 kG) and are also seen to increase rapidly with frequency. This is consistent with the known effects of flux-density and frequency on the core losses of all ferrous materials.

OTHER MATERIALS

An alternative material which is suitable for powder metal technology is Metglas [5]. It is available in the form of tape or in sheets, both of (fine thickness (0.0007 in to 0.0008 in). It can also be pulverized to the powder form required for compacting, where the particle size is as small as 20 microns.

The primary advantage of this material is in the extremely low core losses compared to the other electrical steel alloys. This is a significant advantage in applications where the frequency is high and the need to control the core losses is great. But even at 60 Hz, the core losses of Metglas material 2605-S2 [5] were determined to be only about a quarter of the losses of commercially available silicon steel (10).

The process of making parts with this material is similar to that detailed for other powder metals. There is, however, the method of explosive compaction which is an alternative to the mechanical compaction. Significantly higher densities can be achieved in this method. The process itself is more economical and requires a lot less initial investment since large dies and presses are not required.

REFERENCES

Technical information was supplied by the 2 companies listed in 1 - 5 below:

1. Hoeganaes Corp., Riverton, NJ 08077

2. Micrometals Inc., Anaheim, CA 92807

3. Ultra Fine Powder Technology Inc., Woonsocket, RI 02895

4. Western Sintering Co. Inc., Richland, WA 99352

5. Metglas Products, Parsippany, NJ 07054

6. "Supplement on powder metallurgy", Design News, March 1984

7. Metal Powder Industries Federation, Princeton, NJ 08540

8. Howard I. Sanderow, "Soft Magnetic Materials P/M Applications", Appliance Magazine, April 1979.

9. "Theoretical & Practical Considerations for the P/M Production of Magnetic Parts", Remington Arms Co., Inc., (Dupont), Ilion, NJ 13357

10. "Comparative Analysis of the Magnetic Properties of Oriented Silicon Steels and Metallic Glasses", S. D. Washko, D. R. Fecich & T. H. Shen, IEEE Transactions on Magnetics, Vol. 18, No. 6, Nov. 1982, p. 15.