BJBWJBgt^^m^^ 



'JMJisuwssrMjrKiits-tswiwn 



Alloy Powder Tests 



Although the powdered metal tests proved successful, the specific 

 output was not as high as had been anticipated by having the anodic and 

 cathodic materials in such close contact. Evidently, surface oxides on 

 the metal (present during ball milling) acted as high resistance 

 electrical barriers to the current flow. 



To verify this theory, a magnesium-iron alloy was conceived in 

 which intimate physical and electrical contact could be established 

 between anode and cathode. Through conventional alloying only a very 

 small percentage of iron (<<1% by weight) can be dissolved. This is 

 much less than the 10% that proved optimum for the previous powdered 

 metal tests. A process of mechanically alloying otherwise immiscible 

 metals [13,14] has been established. In this process, magnesium and 

 iron powders are milled in a high energy ball mill. The powder particles 

 are cold-welded together as the balls in the mill collide. Repeated 

 collisions cause particle fracturing and rewelding, and, eventually, a 

 uniform alloy powder is formed. 



Alloy samples were prepared and tested. Powder particle sizes were 

 in two ranges: ''as-produced'' (larger than 100 mesh) and ''selected'' 

 (smaller than 100 mesh). The initial tests showed that the as-produced 

 samples reacted more effectively. An alloy particle would generate a 

 gas bubble, rise to the water surface, release the bubble, and sink. On 

 the other hand, the selected (finer) powder would not release its gas 

 bubble. This resulted in a foam surface that actually lifted the powder 

 particles out of the seawater and prevented them from reacting to comple- 

 tion. Thus, the remaining tests were conducted with the "as-produced'' 

 particle size. 



Figure 14 shows the relationship of milling time to reaction 

 efficiency. A maximum reaction efficiency of approximately 90% is 

 approached asymptotically as shown in Figure 14a. Figure 14b shows the 

 30-minute milled alloy achieves almost this 90% completion in 1 minute. 

 For additional tests the 30-minute milling time was used as the standard. 

 (The rapid reaction completion time is desirable for using the alloy as 

 a fuel.) 



Initial alloy compositions were held constant at 5 atomic percent to 

 show the effects of particle size and milling time. Another group of 

 alloys of various iron percentages was tested to develop a family of 

 curves that relate reaction efficiency and power output to alloy composi- 

 tion. These data are shown in Figures 15 and 16. The zero-percent ball- 

 milled magnesium powder was tried to see if sufficient strain energy was 

 stored in the particles to cause stress corrosion. As can be seen in 

 Figure 15, some small amount of iron is required to produce a usable 

 reaction. Alloys of several alternate cathodic materials were tested to 

 see what reactions they would produce. These results are shown in 



10 



