Flatness:
A lot depends on how thin a first layer you are trying to print. If I stick with something in the 0.2 to 0.3 mm thick first layer, a flatness of 0.02 mm (slightly under .001") would be ideal. Anything over 0.01" will be difficult to use. That would be a total error over the print area compared to what ever I'm calling the Z zero plane. Any bump or pit that's < 0.01" dia probably does not count .
Lapping:
Most of the home friendly approaches to lapping surfaces give you a rounded surface. I don't have the big optical flats here at home that we use at work to correct out those sort of errors. I could probably rig the monochromatic light source and the polarizer. The big slabs of optical glass - not so much.
Calibration:
You might be able to measure the plate with the printer. How you would calibrate out all the printer errors without a flat plate - not real clear. What you care about for *accuracy* then becomes an issue. I would prefer to hold the accuracy down in the 0.005" range if at all possible.
Plate Spec:
I believe that the 0.005" spec on the plate is for the "as shipped" lump of metal. I believe that lump is in the 4' wide by 10' long range. I have a lump, but not that big a lump. Since the piece has been through a lot since it left Alcoa, who knows what shape it's in now.
A lot depends on how thin a first layer you are trying to print. If I stick with something in the 0.2 to 0.3 mm thick first layer, a flatness of 0.02 mm (slightly under .001") would be ideal. Anything over 0.01" will be difficult to use. That would be a total error over the print area compared to what ever I'm calling the Z zero plane. Any bump or pit that's < 0.01" dia probably does not count .
Lapping:
Most of the home friendly approaches to lapping surfaces give you a rounded surface. I don't have the big optical flats here at home that we use at work to correct out those sort of errors. I could probably rig the monochromatic light source and the polarizer. The big slabs of optical glass - not so much.
Calibration:
You might be able to measure the plate with the printer. How you would calibrate out all the printer errors without a flat plate - not real clear. What you care about for *accuracy* then becomes an issue. I would prefer to hold the accuracy down in the 0.005" range if at all possible.
Plate Spec:
I believe that the 0.005" spec on the plate is for the "as shipped" lump of metal. I believe that lump is in the 4' wide by 10' long range. I have a lump, but not that big a lump. Since the piece has been through a lot since it left Alcoa, who knows what shape it's in now.