The Electropolishing of Aluminium 



325 



Fig. 2. Typical curve showing the relationship between cur- 

 rent and voltage in the polishing bath. 



necessary to control the conditions with great care. The 

 factors affecting the polish were investigated in consider- 

 able detail, and this work is being published elsewhere. 

 One or two of the more important results will be men- 

 tioned here. 



It is found that the progress of the polishing process 

 can be followed very well by plotting the voltage applied 

 to the bath against the current flowing through it. A 

 typical polishing curve is shown in fig. 2. In the region 

 AB, the current increases more or less linearly with vol- 

 tage, and no polishing action takes place. At B, an un- 

 stable condition sets in, associated with the formation 

 of a high resistance layer at the anode. When stability 

 is again established at C, relatively large increases in 

 voltage are required to increase the current through the 

 bath, and, in this region CD, the aluminium forming the 

 anode of the bath is polished. 



The effect of temperature change, with other parame- 

 ters constant, is shown in fig. 3. It was found that, even 

 though the obvious variables such as temperature, stir- 

 ring of the electrolyte, and specimen area polished were 

 kept constant, the polishing curves were not always 

 reproducible. The principal reason for this was found to 

 be in the approach to the point of instability B. If the 

 voltage was increased too quickly in the region AB, the 

 voltage sometimes increased above the minimum at 

 which the instability would start. If this occurred, the 

 whole of the subsequent curve was modified. When the 



to 



Fig. 4. Controlled polishing at points on the characteristic 



voltage increments in the region AB were made small, 

 and with adequate time intervals between them, the 

 polishing curves were found to be reproducible. 



Expcrinie/itdl Proccdiiie. -In order to relate the struc- 

 ture resulting from a polish to the polishing conditions, a 

 polishing curve was hrst established for a given size of 

 specimen and for fixed temperature and stirring speed. 

 Following the procedure described abo\e, a scries of 

 similar aluminium specimens were polished under condi- 

 tions defined by the points P^, P^, Pj ... on the portion 

 CD of the curve (Fig. 4). After five minutes polishing at 

 such a point, the specimen was removed from the bath 

 with voltage still applied, washed in water and alcohol, 

 and dried in warm air. 



It was found that thin unbacked films of formvar 

 could be water stripped frotn the aluminium withcnit 

 difficulty, provided that the process was completed within 

 about half an hour of the preparation of the specimen. 

 With greater delays than this, stripping became progres- 

 sively more difficult, and, in some cases, a thick backing 

 film had to be used. The replicas were subsequently 

 shadow-cast. Several portions of each replica film were 

 examined in the electron microscope, and a number of 

 photographs taken at random on each; this ensured a 

 representative selection of available structures. 



When structure appeared on a specimen, the di- 

 mensions (i.e. the separation of the furrows, the 

 spacing of the dots, and the width of the elements 

 of the whorls) were measured. It was found that these 

 dimensions were substantially constant for a given 

 aluminium specimen. Each specimen produced at 

 the points F,, Pt, etc. on the polishing curve was 

 found to have its own characteristic structure spacing. 





Fig. 3. Effects of temperature on polishing curves at constant 

 stirring speed (3000 r.p.m.). 



Fig. 5. Controlled polishes at 0"C, 3000 r.p.m. 



