Just to throw a stick on the fire; you can use the "bleed back" method if you are attempting to establish whether you have excavated deep enough to remove the defect.

The discontinuity contains a volume of penetrant. The volume retained is referred to as the reservoir. A certain amount of the penetrant will exude from the discontinuity and spread through the developer to produce a stain and magnify the apparent size of the discontinuity.

When you attempt to remove the discontinuity by grinding or using a rotary file (burr), you may smear the soft aluminum thereby sealing the opening connecting it to the surface. However, if you don’t smear the metal, the penetrant remaining within the discontinuity will continue to exude and will continue to produce an indication as long as there is penetrant remaining in the reservoir and there is an opening to the test surface.

The bleed back technique is often used to confirm whether the indication is a false indication or a true indication. It can also be used to better characterize the nature of the discontinuity causing the indication. A very large reservoir will produce a brilliant indication as soon as the non aqueous developer has dried. The bleed back technique produces an indication at a slower rate because there is less penetrant in the reservoir after each test.

There is always a concern that metal removal operations will smear soft metal. Whether or not you repeat the entire penetrant test is not going to produce indication (stain) if the opening (discontinuity) to the test surface is smeared and sealed. Many aerospace specifications require an additional etching operation before the penetrant test to ensure any smeared metal is removed.

The problem of smearing soft metal is always a concern when a weld is finished by grinding, machining, or sanding. I will usually tap the surface of a finished (sanded or ground) aluminum weld searching for signs of porosity. Lightly tapping the surface of the weld with a stiff bristled stainless steel brush will produce little dimples if there is any porosity sealed with smeared metal. I “pop” the smeared metal with a pencil point, then perform the penetrant test or send the part out for etching.

My advice is to use the bleed back technique, i.e., wipe the area with a dry clean wipe and reapply the developer. If the defect still has an open path to the test surface the indication will reappear. Exercise some care that the test piece isn’t ground with such force that it becomes hot enough to dry the penetrant remaining in the reservoir. If the indication reappears, continue to excavate metal. If the stain does not reappear, repeat the test, but keep in mind any smeared metal can and will seal the opening, thus no indication will form even if the defect is still there. 

There is more to the certification process than getting x-hours of training and taking a test. Your employer has to engage the services of an Examiner (Level III) that is qualified per NAVSEA TP271 if you are working to NAVSEA TP278. The Examiner has to develop a written practice and penetrant procedures that have to be reviewed and approved by your customer before you can implement them in your shop. Likewise, any laboratory that is providing testing services for your employer has to have all the approved procedures in place before they can perform NDT on your employer's behalf. Your employer should have submitted the laboratory's written practice and penetrant procedures to your customer before awarding them any work. 

Best regards - Al