Not to sure of the necessity for extreme flatness of your sole
plate. However a very convenient visual check is the ancient
plate glass methylated spirits capillary test.
Lay a thick piece of glass on your sole plate with one edge of
the glass propped up on a thin spacer chosen to give an airgap
slope of about 1 in 100.
Carefully introduce a small amount of methylated spirits into
this airgap. Very little is needed - the amount carried in a
small artists brush is sufficient. Capillary action will draw
the spirits to the touching end of the airgap. Add enough for a
band of liquid about 1/2" wide. Surface tension causes the free
edge to have uniform thickness so the height of the free edge is
a direct indication of 100:1 amplification of the airgap at the
Because the meths is flowing in a very small gap it
takes a little time to settle into its final position. The
smaller the slope the greater the amplification and the longer
the settling time.
The flatness quality of the glass can first be checked by
the same method on your surface plate. The glass doesn't need to
be very thick because it's fully supported along its length.
1/4" or more is OK.
The best glass to use is genuine plate glass as this has
been ground flat and polished on both surfaces. This is not too
easy to find.
Most current glass is float glass. This process inherently
produces glass with two precisely parallel surfaces
but the flatness is dependent on the precision of the mechanical
alignment of the outfeed rollers. This may introduce a small
amount of twist or bow. Its quality as a flatness standard is a
bit variable so selection is necessary.
A much better bet is mirror glass (e.g. an ancient dressing
table mirror). Twist or bow in a reflective surface is
unacceptable so quality mirrors are either plate glass or float
glass selected for minimum twist/bow.
The problem is removing the backing. The silvered surface is
protected by a copper plate layer followed by a paint or resin
film. The film can usually be removed by a methyl chloride based
paint stripper but nitric acid is needed to shift the copper and
For precision work with fine ground optical surfaces a 1 in
1000 slope can be used. For coarser work a similar method is
possible but with a 1 in 20 slope and the meths replaced by a
handful of identical smallballs.
The method is still usable with glass that is not dead flat.
If the workpiece is stationary and the glass position is
changed, workpiece errors stay put, glass errors move with the
Then put a blade in it which is usually hand ground and stone finished
and therefore imperfect.
PS you can get a plane with a flexible sheet base which can be adjusted
concave or convex what do you do then?
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.