# measuring a plane sole flatness?

I have a passing interest (some might say obsession :-) in making woodworking plane soles flat.
(aside; they clearly need to be "fairly flat";
no-one is quite sure "how flat" they need to be; I take the view that "too flat" is "flat enough" ;-)
http://www.geocities.com/plybench/flatten.html
I have (for a while) had a Chinese surface plate, like this one:
http://www.leevalley.com/wood/page.aspx?sid=&ccurrency=3&page2526&category=1,43513
I now have a Mercer G.303 dial indicator (second hand)
http://www.cromwell.co.uk/MER3003187V
And 2 less accurate dial indicators, one with long, linear travel, but only 1/2 thou resolution,
I also own some surface gauge stands, both branded and apprentice made.
My workshop is a (messy) mixture of woodworking and metalworking stuff.
http://i48.photobucket.com/albums/f234/bugbear33/workshop.jpg
How can I go about measuring the shape of my (hoped to be flat) plane sole as accurately as possible?
BugBear
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Nick Mueller wrote:

So the surface gauge is sitting on the surface being measured - no requirment for a surface plate?
I'm not getting the picture here - since a (e.g.) #5 plane is 14" by around 2 1/2" I don't see how anything gets turned through 90 degrees, but since my visualisation of what you're describing is so poor, that may be natural!
Can you (for this ignorant person) be more explicit about where the surface gauge base is, and what point is being contacted by the dial gauge, and how the various dial gauge reading can be translated/interpreted into a quantative maps of the bumps and hollows?

When you say ruler, do you mean cast iron straightedge, wide enough for things to sit on?
http://www.indiamart.com/minimachinetools/pcat-gifs/products-small/cast-iron-camel-back-straig.jpg
BugBear
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
bugbear wrote:

Yes!
OK, its a "bit" on the long side. So the method I tried to describe in poor wording might not be the best. :-))

OK, new game, new luck. Put the stand in one corner. Adjust the arm to reach out a bit less than the sole's width (2 1/2"). Now the dial's tip sits on the opposing corner where the stand is. Adjust to zero. Rotate stand 90°. You know what direction to rotate? :-)
Now you get a difference. It shows you almost (not completely, but quite good) wether the sole is bent upward or downward. Move the stand along the longer side in such a way, that if you rotate the stand, the dial's tip reaches the place where the stand was standing before. Read the dial, rotate 90° (reaching the opposing side), read, rotate further 90°, read, move stand, ... Do that all along the sides of the sole. Now you did get some kind of map. Try to figure out how the sole has to be warped and bent to get the readings. That method is a bit complicated and is better to *verify* flatness.

I should have said straightedge? Some precision ground flat steel. Not the ones with a knife that you use against the light. You get them in different grades. I do have one 1m long in grade 0 (actual deviation 0,007mm). A 500mm straightedge grade 1 should be by far good enough for you and would put you back at about 50.- EUR. Unfortunately, they are very bad at using them with touching blue. But very good at checking straightness *along* *a* *line*. You can always use that straightedge on the diagonals to show twist. Sorry, I don't know the right word in English for that "ruler".
HTH somehow. :-) Nick
--
The lowcost-DRO:
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Nick Mueller wrote:

Ah; each measurement is measuring a local gradient, and these can be integrated to reveal a curve. I understand.

A ruler is "ruled" i.e. it has divisions in inches, or millimeters.
A straight edge doesn't. Good rulers may be used as straightedges.
Straightedges come in two main styles; "ground flat stock"
http://www.bowers.co.uk/products/show/577
and the ones which are really long, narrow surface plates, either camel back cast iron:
http://www.gandmtools.co.uk/cat_leaf.php?id284
or granite blocks:
http://www.tru-stone.com/pages/smp.asp
I can see how to use that for simple checking, but I don't understand hwo to use it as you said in your earlier post:

I admit, I had envisaged setting up the plane (upside down) to the sole was parallel to the surface plate, and using a dial indicator from above, with the indicator support sliding over the surface plate.
BugBear
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
wrote:

If you make up a little block - say 1"x 1" and drill and tap one face with 3 screws in a triangle to another face at 90 degerees to the first, you fix a piece of front silvered miror. Now turn the plane over and mark it out in pencil in 1" squares, next put the block in one square standing on the screws and shine a builders lazer at the mirror from about 10ft and plot the reflection from the lazer beam on a piece of card by the lazer, so the light has travelled 20ft. You will be able to easily calculate the error in each square. For grearer accuracy repeat with the mirror standing at the junction of each square. I hope you find something more interesting to do over christmas Peter
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

I say your going over the top ........for a stanley plane
the plane may end up perfect .but wood will never be ........swelling and changing with the moisture content within hours of you getting it bang on .
if you want to go to the trouble.. do it with a plane that's going to be worth a lot of money at the end of all your hard work.
stephen thomas in the practical machinist forum builds bespoke wood planes..........search his posts ..
all the best.markj
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

I maybe missing something, but why dont you use the traditional method of bluing and then scraping the high spots? It will get you as flat as the reference with a little patience. I assume you dont really care about the mapping of the surface, just that it is flat?
Dave
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Rest it sole down on the surface plate and see where you can get a feeler gauge under it.
Or, if it is very good, rest on cigarette papers and see which ones are touching and which not.
Or blue it up and and see where the high spots are.
--
Charles Lamont

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

are
Or bung it on a surface grinder (or even decent flat belt sander) and skim to a clean up cut, and accept it'll be perfectly adequate for the job thus freeing up time to worry about things worth worrying about <G>
AWEM
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
On Mon, 22 Dec 2008 22:43:35 -0000, "Andrew Mawson"

I've never felt the need to measure a sole for flatness, or even bung it on the surface grinder for that matter. A couple of minutes in the frying pan, then a squeeze of lemon juice is all that it needs <G>
Regards, Tony
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

a
ones
and
the
bung
the
That's because you don't have a soul - we all know that <G>
AWEM
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
On Tue, 23 Dec 2008 08:11:48 -0000, "Andrew Mawson"

Ah - but I do have a kipper <G>
Regards, Tony
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Charles Lamont wrote:

I've done that. My site that I linked to explains how to do it.
I wanted to measure, quantitively, how good (or bad) a job I'd done.
To those telling me not to bother, I would point out that most model engineering is done for the sake of the achievement, or enjoyment of the task itself, not because the final item is intrinsically "needed".
I am also well aware of the debate concerning how flat a plane's sole needs to be, and also how much effort it's worth putting into a low-value second hand tool, and what the economic return might be.
BugBear
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Your measuring method needs to be an order of magnitude more accurate than the thing you are measuring. You need a Talysurf.
--
Charles Lamont

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

To be theoretically correct, shouldn't the back and front of the sole be offset from planar at the blade by the depth of cut thickness? I think planing machines actually do this, since they have significant depths of cut. You might also want to evaluate the change from flat due to the frictional heat of the planing and the bending pressure exerted on the handle(s). ;>)
Don Young (USA)
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
wrote:

If you are going to that kind of extreme for what is, at the end of the day, a pretty crude a hand tool, then you might also want to get a life ;-)
Regards, Tony
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Tony Jeffree wrote:

This *is* the forum for people who make 1/5 scale bicycles, yes?
BugBear
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Could it just be that your irony detector was switched off?
David
--
David Littlewood

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Don Young wrote:

Indeed. On a mailing list, someone proposed the "manx jointer", a hand plane (on this handtool only list, "tail" is a euphemism for power lead, and manx cats have no tail) that had the sole-heel step you propose.

Discussed here: