It's been a slow day on the oldtools list, so I had to fill out
   my dinner break with _something_... :-)

   So far we have a concrete estimate by the ISO of acceptable standards
   for plane sole flatness.  As far as I know, the derivation of that
   standard did not take into account sole flexure, i.e. bending of the
   sole of the plane under downward pressure applied to the plane,
   through the tote, while planing.

   It's not hard to get a reasonable idea of what effect that would
   have, for a cast iron bench plane, using standard structural
   analysis foo.  The exact analysis allowing for precise shape and
   so forth is tedious, but we can get a reasonable ballpark figure.

   So, consider a Stanley bench plane.  The principal elements to
   consider are the sole ( natch, ) and the sides.  The other stuff
   is more or less just floppin' around in there.  We'll assume the
   sole is concave from end to end by some measure, and that during
   use to sole is initially supported at the front and back ends,
   and moreover that the downwards force applied through the tote
   is centrally located, i.e. roughly in a spot equidistant from the
   front and back edges.  I'm thinking in particular of a Stanley #8,
   which will likely be the most extreme case where we could get
   flexure, and will constitute an upper limit of the amount of
   flexure we could get, for equivalent dimensions.

   All figures for cast iron.

   The Young's modulus, or modulus of elasticity, of cast iron, is:
      E = 152.3 GPa = 22.1x10^6 psi

   The rigidity modulus is similarly:
      G =  60.0 GPa =  8.7x10^6 psi

   The deflection of the sole plate alone will be:

      d(sole)  =  (WL)/(bh) * ( L^2 / (4E h^2 ) + 3/(10G) )

   and the deflection of the sides alone will be:

      d(sides) = (2WL)/(BH) * ( L^2 / (4E H^2 ) + 3/(10G) )

   where:

      W = force applied, in pounds.
      L = length of plane, in inches.
      b = width of sole, in inches.
      h = thickness of sole, in inches.
      B = thickness of a side, in inches.
     
   Also, H = a "typical height" of the side of the plane.  One might
   choose a value for this which is equal to the minimum value of the
   side height across the length of the plane.  OTOH, one might choose
   the maximum side height, which wouldn't be very realistic.  We'll
   compromise and define this to be just a little more than the 
   average side height out where it's not so high - we want to see
   what's the maximal flexure we can get realistically.

   Plugging in the values for E and G, we get:

      d(sole)  =  (WL)/(bh) * ( 1.13x10^-8 (L/h)^2 + 3.45x10^-8 ) inches,

      d(sides) = (2WL)/(BH) * ( 1.13x10^-8 (L/H)^2 + 3.45x10^-8 ) inches.

   Now consider our Stanley #8, for which we have the following data:

      L =   24 inches,
      b =    3 inches,
      h = 3/16 inches,
      B =  1/8 inches,
      H =    1 inch.

   Plugging in that data, we arrive at:

      d(sole)  = ( 7.9x10-3 W ) inches,  or 7.9W thou.
      d(sides) = ( 2.5x10-3 W ) inches,  or 2.5W thou.

   Remember that W is numerically equal to the downwards force in
   pounds, applied to the plane in use, i.e. whenever this _matters_
   for want of a better way of putting it.

   The sole and sides operate in unison, rather than independently,
   of course, so I'd tend to think we can safely estimate the total
   sole flexure at around 2W thou.

   So, we'd be led to believe that a downwards force of 10 pounds
   during planing, not totally unrealistic, would result in a
   downwards mouth deflection of around 20 thou, or around 1/64 inch.

   Make of this what you will.  It always has to be taken into
   account in considerations of sole flatness and its significance.

   For metal bench planes, this is likely gonna be a maximal figure,
   i.e. I doubt it would get any higher.  The #8 is a _big_ plane,
   and it's _long_.  OTOH, it's pretty sturdily built.  Nevertheless,
   shorter planes will have the point of applied pressure more and
   more further back from the centre point of the plane.  E.g., for
   a #3 or #4, you get relatively little pressure in the middle of
   the plane counteracting the natural concavity of the sole, and
   flexure there will be _significantly_ less, much less so than
   the above numbers and formulas would indicate methinks.

   I've given the formulas above, and if it amuses you, you can
   plug in numbers for other planes, or modify the reasoning at
   your will.  Again I emphasize, this is just a back-of-the-envelope
   calculation there, but anyone here is welcome to make it more
   precise, or certainly check my figures ( I was never very good
   at sums, let alone noticing typos. )

   The generalization to wooden bench planes, or more exotica, is
   left as an excercise for the reader.

   Doug Dawson
   dawson@p...

>
>   So, we'd be led to believe that a downwards force of 10 pounds
>   during planing, not totally unrealistic, would result in a
>   downwards mouth deflection of around 20 thou, or around 1/64 inch.
>

assuming your calcaulations and reasoning is correct (I'm far to lazy to
check) it seems like you've draw the conclusion that sole flatness just
doesn't matter.  after all, no one I know, least of all me, would use a
plane that showed a 64th of daylight under the mouth (*that* lazy I'm not)
and that normal user-pressure would assure the mouth is snug up against the
work.

but then you point out that user-pressure won't deflect a smoother quite so
much.  true.  and also we're pretty much disregarding the fact the sole may
be *twisted* instead of simply concave, front-to-back.  I think the way the
plane is constructed, you've got a heavy frog tending to resist any user
pressure in the short axis of the plane (but also trying to flatten it in
that axis as well.)  this is all getting complicated.

back to the original example.  okay, you've got an gross concave surface
and user pressure holds the center flat to the work.  but how much pressure
is exerted against the work *at the mouth* after this deflection is
flattened out?  in the example given: 0.  in effect, the mouth has *no*
effect.

anyway, personally, I don't have any planes this far out of tune, so I
can't speak to how well they do or don't perform.  I have had planes with
unflat bottoms and I've tuned (flattened) them and they seem to work
better.  I haven't done this with a #8, but I've done if with a couple of
modern blocks and have found them vastly improved.

415-267-7313(beep/Bay Area) 310-201-8615(beep/Westwood)

Doug Dawson did some marvelous estimates regarding plane sole flexion (I'm 
going to assume that's a real word), based on certain assumptions:

>   We'll assume the
>   sole is concave from end to end by some measure, and that during
>   use to sole is initially supported at the front and back ends,
>   and moreover that the downwards force applied through the tote
>   is centrally located, i.e. roughly in a spot equidistant from the
>   front and back edges.

The only problem is, how dows one apply force *downwards* at the iron, for 
purposes of determining sole flexion, when the hands are located at the 
front ball and the tote, which are some distance away?  Even if one pressed 
down like the devil with both hands, that wouldn't suppress the concavity, 
if any, halfway between the places where downward pressure is applied.  The 
downward force applied to the tote, for example, would be felt, at most, at 
the bottom of the tote, so the principal flexion would occur between the 
bottom of the tote and the back end of the plane, and similarly for the 
downward force on the ball.  Assuming that one applies sufficient force to 
drive the sole flat to the surface at these two points, the concavity would 
remain, to some lesser degree, between these pressure points (i.e., around 
the mouth).  Ain't nobody pressing down on the iron.  Of course, the depth 
of cut of the iron will also result in downward force, flattening the sole 
further.  I suspect that any useful modeling of the forces involved in this 
situation will call for higher mathematics.

----------------------------------------------------------------------
Michael D. Sullivan, Bethesda, Maryland, USA
mds@a... / avogadro@w... / 74160.1134@c...
----------------------------------------------------------------------

Doug Dawson wrote:
 
~     It's been a slow day on the oldtools list, so I had to fill out
~     my dinner break with _something_... :-)
~  
~     So far we have a concrete estimate by the ISO of acceptable standards
~     for plane sole flatness.  As far as I know, the derivation of that
~     standard did not take into account sole flexure, i.e. bending of the
~     sole of the plane under downward pressure applied to the plane,
~     through the tote, while planing.

Now this is something new! Standards as estimates. How does one have a
concrete estimate? 8-).
  
~     It's not hard to get a reasonable idea of what effect that would
~     have, for a cast iron bench plane, using standard structural
~     analysis foo.  The exact analysis allowing for precise shape and
~     so forth is tedious, but we can get a reasonable ballpark figure.

What is a "Ballpark" please? - somewhere where this strange American
game of rounders is played perhaps, (the game being strange of course,
never the Americans, never, oh dear me, never!) and what have figures
got to do with it? Do they never know the exact score or something?
8-).

.

~     Now consider our Stanley #8, for which we have the following data:
 
~     Plugging in that data, we arrive at:
~  
~        d(sole)  = ( 7.9x10-3 W ) inches,  or 7.9W thou.
~        d(sides) = ( 2.5x10-3 W ) inches,  or 2.5W thou.

~     Remember that W is numerically equal to the downwards force in
~     pounds, applied to the plane in use, i.e. whenever this _matters_
~     for want of a better way of putting it.

~     The sole and sides operate in unison, rather than independently,
~     of course, so I'd tend to think we can safely estimate the total
~     sole flexure at around 2W thou

~     So, we'd be led to believe that a downwards force of 10 pounds
~     during planing, not totally unrealistic, would result in a
~     downwards mouth deflection of around 20 thou, or around 1/64 inch.

Messing about, plane in hand, with bathroom scales on the bench top
(to examine another theory), I found that with my puny 11 stone (154
pounds) I can register about 90 lbs on the scale. According to Doug's
figures, this makes a deflection of about 180thou or 0.18 inches. Cor
blimey! Cor luv a duck!

Perhaps we should take off our hats, (or maybe some other garment?) to
Doug for so nobly sacrificing his dinner break for our enlightenment.
 
~     Make of this what you will.

Done that Doug! 8-).

~     ..........................   It always has to be taken into
~     account in considerations of sole flatness and its significance.

Of course we should pay tribute to this static analysis. What I wonder
would be the outcome of an analysis of the dynamic situation, taking
into account of the need progressively to adjust (no split infinitives
here, no siree) for temperature rises. To what extent does suction on
non-corrugated soles affect this deflection, one gently enquires?
 
~     ....... counteracting the natural concavity of the sole, and
~     flexure there will be _significantly_ less, much less so than
~     the above numbers and formulas would indicate, methinks.

However, from where does the ISO's Chief Savant derive this concept of
/natural/ concavity (undefined in extent) of a plane sole? Observation
and testing of a statistically significant quantity? Workshop lore?
Exchange of data at one of these swapmeets one reads about? Mind you,
most of us from time to time can get preoccupied with certain
concavities, and convexities for that matter, especially when
well-dined on a Sunday evening.

Perhaps next time Doug is well-dined, he might favour us with his
formula for natural concavity, no doubt including factors such as age,
(the plane not the owner!) nature, location and area of patent marks,
length, presence of a complete decal (whatever that is), whether it
has a low or high knob, factors for rosewood and otherwise, japanning
percentage, whether bedrock or not and so on. Minor factors such as
mis-match of knob and tote patterns can perhaps indulgently be
excused.

Jeff 

-- 
Jeff Gorman - West Yorkshire
jeff@m...

I just went downstairs and measured the flexibility of three planes.

Here is the data, someone else can compute if there's any of that
to be done...

All three planes were supported 1" in from their extremities and
a 10 lb weightlifting weight was placed vertically and straddling
the plane right over the mouth.  The deflection was read off the 
sole, in the center, about 1/8" in front of the mouth.

                     distance                
    plane         between supports          deflection

  old    #8            21-3/4"                 0.0020
  2 pd   #605-1/2      13"                     0.0005
  CDN SW #5            12"                     0.0008

Paul

Montreal (Quebec)

   Paul wrote,
 
> I just went downstairs and measured the flexibility of three planes.
> 
> Here is the data, someone else can compute if there's any of that
> to be done...
> 
> All three planes were supported 1" in from their extremities and
> a 10 lb weightlifting weight was placed vertically and straddling
> the plane right over the mouth.  The deflection was read off the 
> sole, in the center, about 1/8" in front of the mouth.
> 
>                      distance                
>     plane         between supports          deflection
> 
>   old    #8            21-3/4"                 0.0020
>   2 pd   #605-1/2      13"                     0.0005
>   CDN SW #5            12"                     0.0008

   I changed the support points in like that from 24", adjusted
   the loading and measuring points in to where you put them, all
   only for my Stanley #8, and got a deflection of 0.007 inches.

   Thanks for measuring those things, this is gonna get more
   interesting as I keep working on it ( in my spare moments when
   I'm exhausted for anything else. )  Yep, there's gonna be more
   parts...  this stuff is too accessible to be the least bit
   controversial.

   My estimates were geared to be maximal estimates, i.e. I tried
   to give the people who say planes flex like crazy as much benefit
   of a doubt as possible.  I.e., I'd feel safe saying that you
   wouldn't see any _more_ deflection than that, pretty much.  To
   that end, I had set the side height to be roughly it's minimal
   average value of an inch.  The side height of mine peaks at 
   2 3/8 inches, and is above 1.5 inches for a good 5 inches of
   plane length.  If you set the "effective side height" to 1.5
   inches, my earlier calculation matches your result of 0.002 inches
   exactly.  I think a good estimate of the effective side height is
   one of the most critical things here, as far as that calculation
   went, barring a more elaborate analysis.

   Stay tuned for Parts III+, and notes on the earlier replies to
   this thread.

   Doug Dawson
   dawson@p...

Doug D. wrote:

(snipping much good analysis of what's going on with flexure...)

I really hesitate to get out out of my quiet, comfortable spot in the corner
next to the cracker barrel.  I've been there stuffing my mouth full of
crackers to prevent any more too-quick comments.

But, Doug, phtooie, phtooie!....(oops, sorry, I'll clean it up that mess in
a minute...oh, are those your good pants?)

>   It's been a slow day on the oldtools list, so I had to fill out
>   my dinner break with _something_... :-)
Must have been an extraordinarily tiring one.

I really don't want to rekindle any long smouldering differences  between
theorists at their desks and engineers who have to actually face the "real
world", so don't take this wrong.  For the sake of those who might have
copied your plane sole deflection calculations onto the front of the
refrigerator, I gotta say, I think you missed (due to fatigue) some
important things that really mess up the results.

#1 - The interaction between the sides of the plane casting and the sole are
totally different than the way your math treats them, since they are
actually one piece.

#2 - I know that for the sake of the KIS principle you didn't deal with the
fact that the side thickness isn't XX" as you said, and is tapered - joined
to the sole with a rather large radius...Well, that alone throws the
calculation way off.  But not as much as #1.

#3 - My Phtooie alarm went off as soon as I read:
>   The Young's modulus, or modulus of elasticity, of cast iron, is:
>      E = 152.3 GPa = 22.1x10^6 psi
Among REAL WORLD practical people, it has become pretty standard to apply a
FF (Fudge Factor) to anything that comes from theoretical physics.

To be more constructive - 
A factor your formula needs is something I picked up at a public library
several years ago out of a 40 year old book.  It's a table of factors to be
used as divisors when using Young's modulus here on the surface of this REAL
world, where I spend most of MY time. 

Reginold T. Smith (of Austrailia somewhere) created a fudge factor table to
deal with the discrepancies I'm talking about.  (He even labeled the "units"
REGS  - must have been an ego thing)  

Well, I had copied it down, and had played with it before, so I dug it up
now, just to check out your theoretical  results. Because they just didn't
sound quite right to me.

The revised (your) formula for a piece 24" long says 10lbs. at the mouth can
deflect the surface .003.  Kind of a long way from what you suggested I think.

So (since I don't have cable TV) I checked it out on a #8.  Sure enough!
(although admittedly I got .004" with 10 lbs.)

But, going back to specific refinements needed for your formula -

#5 -  I calculate that the force directly downward on the knob and tote to
produce 10 lbs at the mouth at 18.6 lbs.  Ignoring the problem of doing that
with the shape of the tote, what it would take to deflect it .020" would
leave my feet off the floor.  How am I going to give it a push forward?

I'll reprint Mr. Smith's conversion table I suppose, for anyone really
interested.  But it's even longer than this post.

Long note just to say "Back to the drawing board, Doug."  And I hope this
week isn't as exhausting for you.  :^)

------
Gene

> Doug D. wrote:
>
> (snipping much good analysis of what's going on with flexure...)
>
> I really hesitate to get out out of my quiet, comfortable spot in the
> corner next to the cracker barrel. I've been there stuffing my mouth
> full of crackers to prevent any more too-quick comments.
>
> But, Doug, phtooie, phtooie!....(oops, sorry, I'll clean it up that
> mess in a minute...oh, are those your good pants?)

   S'okay, their only my work skins.

> >   It's been a slow day on the oldtools list, so I had to fill out my
> >   dinner break with _something_... :-)
> Must have been an extraordinarily tiring one.
>
> I really don't want to rekindle any long smouldering differences
> between theorists at their desks and engineers who have to actually
> face the "real world", so don't take this wrong. For the sake of those
> who might have copied your plane sole deflection calculations onto the
> front of the refrigerator, I gotta say, I think you missed (due to
> fatigue) some important things that really mess up the results.
>
> #1 - The interaction between the sides of the plane casting and the
> #sole are
> totally different than the way your math treats them, since they are
> actually one piece.

   Absolutely right, this has to be taken into account. OTOH, if you
   look at the relative figures, the sides are dominant in resisting
   flexure, the sole is only a perturbation on that.

> #2 - I know that for the sake of the KIS principle you didn't deal
> #with the
> fact that the side thickness isn't XX" as you said, and is tapered -
> joined to the sole with a rather large radius...Well, that alone
> throws the calculation way off. But not as much as #1.

   OOPS!!! _Most_ excellent point. It approaches 3/16" at the base. That
   increases the effective side height even further, even moreso at the
   centre hump ( - see my reply to Paul. )

> #3 - My Phtooie alarm went off as soon as I read:
> >   The Young's modulus, or modulus of elasticity, of cast iron, is: E
> >   = 152.3 GPa = 22.1x10^6 psi
> Among REAL WORLD practical people, it has become pretty standard to
> apply a FF (Fudge Factor) to anything that comes from theoretical
> physics.

   That came from an engineering handbook. I'm not really sure what the
   tolerances were, because they didn't quote them. ( *$%(@&^(*
   engineers, I've lost track of the number of courses in experimental
   statistics I've taught them, and they just never learn the stuff,
   it's not in their nature. :-( Pretty typical - I hope they learn THAT
   in the real world, cuz they refuse to learn it here. ;-) )

> To be more constructive - A factor your formula needs is something I
> picked up at a public library several years ago out of a 40 year old
> book. It's a table of factors to be used as divisors when using
> Young's modulus here on the surface of this REAL world, where I spend
> most of MY time.
>
> Reginold T. Smith (of Austrailia somewhere) created a fudge factor
> table to deal with the discrepancies I'm talking about. (He even
> labeled the "units" REGS - must have been an ego thing)

   I'd be interested in that. Prolly too specialized to post publicly
   though, unless someone protests. I don't have any resources of that
   sort, and wouldn't have a clear idea of where to look for them, aside
   from generalities.

> Well, I had copied it down, and had played with it before, so I dug it
> up now, just to check out your theoretical results. Because they just
> didn't sound quite right to me.
>
> The revised (your) formula for a piece 24" long says 10lbs. at the
> mouth can deflect the surface .003. Kind of a long way from what you
> suggested I think.

   I'd have to verify the circumstances, doctrine, etc., because in
   something truly quantitative like this you have to watch exactly what
   procedure is being used.

> So (since I don't have cable TV) I checked it out on a #8. Sure
> enough! (although admittedly I got .004" with 10 lbs.)

   Again, precise layout? As noted in my response to Paul, it really
   makes a big difference to the outcome.

> But, going back to specific refinements needed for your formula -
>
> #5 - I calculate that the force directly downward on the knob and
> #tote to
> produce 10 lbs at the mouth at 18.6 lbs. Ignoring the problem of doing
> that with the shape of the tote, what it would take to deflect it
> .020" would leave my feet off the floor. How am I going to give it a
> push forward?

   I was meaning to comment on this in a reply to someone else. With the
   #8, in particular _my_ #8, which looks like any other so I feel
   pretty safe here, pressure applied to the tote during planing would
   present a downwards force virtually equidistant from both ends. And
   that's only the vector component directly into the wood. I said it
   was not totally unrealistic, playing the optimist, but I'll leave it
   to others to comment more carefully on just to what extent that's
   true. I envy the other people who have the measuring equipment to be
   able to determine that, I don't have the stuff. Wish I did.

> I'll reprint Mr. Smith's conversion table I suppose, for anyone really
> interested. But it's even longer than this post.
>
> Long note just to say "Back to the drawing board, Doug." And I hope
> this week isn't as exhausting for you. :^)

   Thanks for the point on the tapered sides. Now Paul's measurement is
   completely consistent with my prediction, at least insofar as what I
   did was meant, as stated, to be a ballpark estimate of how much
   deflection you could likely never _exceed_ reasonably, to well within
   an order of magnitude. All the above said, our results seem to be
   fairly close.

   Again, just a back-of-the-envelope calculation to suggest an extremum
   of possible deflections. You'd really have to do a more elaborate
   analysis, methinks, to improve on it.

   More later.

   Doug Dawson dawson@p...

 Regarding the excellent debate on rigidity of cast iron being 'fathered'
 by Doug Dawson.....

 
 > deflect the surface .003.
 ....
 >.. checked it out on a #8.  Sure enough!  .. I got .004" with 10
 >lbs.)

 > so I feel pretty safe here, pressure applied to the tote during
 > planing would present a downwards force virtually equidistant from
 >   both ends.  And that's only the vector component directly into > the
 wood

 Q1)        Can I take it that, when using a No 8, I can bend it by 3 - 4
            thou?  What happens when, at the final stroke, I let the plane
            caress the surface, doing my best to apply only forwards
            pressure.?

 Q2)        Can I scale the result down for shorter planes?  a No 4, for
            instance will apparently stay as near as dammit to being flat.
            Which is what I want my smoother to be.

 (Lumpy solers disregard last sentence)

 Richard Wilson
 who is lost in admiration at the content of this thread.

 
   Richard Wilson wrote,
 
>  Regarding the excellent debate on rigidity of cast iron being 'fathered'
>  by Doug Dawson.....
> 
>  
>  > deflect the surface .003.
>  ....
>  >.. checked it out on a #8.  Sure enough!  .. I got .004" with 10
>  >lbs.)
> 
>  > so I feel pretty safe here, pressure applied to the tote during
>  > planing would present a downwards force virtually equidistant from
>  > both ends.  And that's only the vector component directly into > the
>  wood
> 
>  Q1)  Can I take it that, when using a No 8, I can bend it by 3 - 4
>       thou?

   It seems like you can, i.e. that's what people are measuring,
   more or less, _provided_ you apply the requisite amount of
   downward pressure, which was ten pounds in that case.  I chose
   that figure, because I kind of doubt that people would ever be
   applying _more_ than that, _downwards_, during normal planing.
   Remembering that we're usually applying force to the plane
   during planing, in a downwards/forwards direction, the
   downwards force will typically ( methinks ) be no more than
   half the total applied force.  Think of pushing a 30 pound
   weight around for some length of time...  It's semi-clear to
   me that that's the upper limit of what we'd be willing to
   tolerate for an supposedly functioning plane.  But I could be
   wrong!  I.e., I don't have the measuring apparatus to determine
   how much force I'm applying to a plane in use.  Anybody willing
   to do some experiments on this, by all means we'd be grateful
   for you're help.

>       What happens when, at the final stroke, I let the plane
>       caress the surface, doing my best to apply only forwards
>       pressure.?
 
   We're working up to that...  but flexure is not gonna be a
   primary factor in what happens in that specific case, it would
   seem.
 
>  Q2)  Can I scale the result down for shorter planes?  a No 4, for
>       instance will apparently stay as near as dammit to being flat.
>       Which is what I want my smoother to be.

   No scaling at this point, for the shorter planes.  The results
   were for planes with a centrally located tote, rather than 
   something with an end-mounted tote like on the #4.  That still
   has to be worked out.
 
>  (Lumpy solers disregard last sentence)
> 
>  Richard Wilson
>  who is lost in admiration at the content of this thread.

   Just say, lumpysolers, flatsolers, it doesn't matter, if all
   we're interested in is the truth.

   Doug Dawson
   dawson@p...

> Doug Dawson wrote:
>  
>    It seems like you can, i.e. that's what people are measuring,
>    more or less, _provided_ you apply the requisite amount of
>    downward pressure, which was ten pounds in that case.  I chose
>    that figure, because I kind of doubt that people would ever be
>    applying _more_ than that, _downwards_, during normal planing.
>    Remembering that we're usually applying force to the plane
>    during planing, in a downwards/forwards direction, the
>    downwards force will typically ( methinks ) be no more than
>    half the total applied force.  Think of pushing a 30 pound
>    weight around for some length of time...  It's semi-clear to
>    me that that's the upper limit of what we'd be willing to
>    tolerate for an supposedly functioning plane.  But I could be
>    wrong!  I.e., I don't have the measuring apparatus to determine
>    how much force I'm applying to a plane in use.  Anybody willing
>    to do some experiments on this, by all means we'd be grateful
>    for you're help.

I was using a little hand held force gage just yesterday at work.  I
playing with a prototype fixture I'd built for a hand assembly line and
had borrowed it from one of the Industrial Engineers to check the
ergonomics.  Turns out it takes about 12.5 lbs to insert the part in
my fixture, so at the moment I have a good feel for what 12.5 lbs feels like.
It ain't much.  Based on my experiments yesterday with this device I 
wouldn't be the least bit surprised if the downward component of planing
was at least ten pounds.  In fact, some of the beefier galoots probably
have hands that weigh more than 10 lbs.  :-)

Keep in mind that all these theoretical calculations don't amount to
anything if the plane isn't concave.  All bets are off if the plane is
flat and is supported by the wood along it's entire length.  (Sheesh,
next thing you know I'll be blowing the dust off my Timenshenko strength
of materials book).

It's a fun discussion, but if I wanted my tabletops to be planed flat within
0.001" I wouldn't have gotten into oldtools.  :-)

-- 
Scott Post   spost@n...
Just say we're exceeding the anal retentive stage here...

Richard W wrote:
> >.. checked it out on a #8.  Sure enough!  .. I got .004" with 10
> >lbs.)
Since that was either my sentence quoted, or matched my results, I wanta reply.
 
> Q1)        Can I take it that, when using a No 8, I can bend it by 3 - 4
>            thou?  What happens when, at the final stroke, I let the plane
>            caress the surface, doing my best to apply only forwards
>            pressure.?
A1) No. I'd be amazed if you have the weight to hold a #8 in the normal user
way and produce 10# direct downward force at the mouth, while managing to
move it in any other direction.  I can't.  Also, my number came from a setup
that would simulate a sole so concave that only the extreme toe and heel are
in contact with wood.  Not a real life situation.

A1.1) >What happens when, at the final stroke, I let the plane
> caress the surface, doing my best to apply only forwards pressure.?
Then the Max 10# becomes only the weight of the unsupported portion of the
plane body, whatever that might be on yours.  And for working wood,
deflection of flatness of the sole becomes a pointless discussion.

> Q2)        Can I scale the result down for shorter planes?  a No 4, for...
>
A2) Yes, very roughly maybe.  Doug built into his formula  the length of the
plane and (I think) it would only go totally wrong when the ratio of
distance from tote to mouth changes significantly.  But as the plane gets
shorter, the relative height of the sides  changes drastically too.  So
stiffness increases.
Stiffness is the only reason they bothered to put them there - right?

A3) Before worrying about this *at all*, be sure you have read Doug's post
on a correction of a numerator/denominator swap typo on the original
formula.  (I didn't see it until he told me).  AND his repeated comments
saying this was a ballpark look at *maximum* deflection normally possible
(not his words - sorry if I said that wrong, Doug)

A4) Something that hasn't come up:  When you look at the shape of the tote,
and the way your hand grips it, a huge %age of the force you can produce
there has to be directed forward.  Much less is directed straight down.
Unless you hold the tote a lot differently than I do.  If you *can* produce
that much YMBAG, and you should work on sharpening that iron.

> who is lost in admiration at the content of this thread.
>
A5) (Taking that as a question without the punctuation mark...)  No one but
me I suspect.
------
Gene

Gene, answering Richard, wrote :

>A4) Something that hasn't come up:  When you look at the shape of the tote,
>and the way your hand grips it, a huge %age of the force you can produce
>there has to be directed forward.  Much less is directed straight down.
>Unless you hold the tote a lot differently than I do.  If you *can* produce
>that much YMBAG, and you should work on sharpening that iron.

Speaking of the iron, isn't the iron trying pretty hard to pull
the plane through the wood at a 45 degree angle ?  Wouldn't 
something like 50% of forward force be converted into downward 
force, and right at the mouth ?

Paul

Montreal (Quebec)

   Paddy wrote,
 
> [tamp]
> 
> It's good to see the members of the flat-sole society having their tea out
> on the front porch today (not that I believe any of this stuff) :^).

   Who among us here is a member of the flat-sole society?  Who said
   anyone here was recommending it?  Not I.  

   Unserviceable villainry! ;^)

   Why, I proudly proclaim myself to be the GREAT SATAN of plowing
   huge gaping chunks out of my surfaces, just cuz I like how they
   look and feel when their lumpy. :-D  

> Paddy GM/ENB/Owns no optically-ground planes.

   I'll get to that...

   Doug Dawson
   dawson@p...

 
    [ I originally sent this Monday afternoon, but it hasn't shown
      up as of this afternoon, so I'm bouncing it back in, as I'd made
      references to it earlier.  Apologies if it shows up twice.  djd ]
 
 
    Paul wrote,
  
> I just went downstairs and measured the flexibility of three planes.
> 
> Here is the data, someone else can compute if there's any of that
> to be done...
> 
> All three planes were supported 1" in from their extremities and
> a 10 lb weightlifting weight was placed vertically and straddling
> the plane right over the mouth.  The deflection was read off the 
> sole, in the center, about 1/8" in front of the mouth.
> 
>                      distance                
>     plane         between supports          deflection
> 
>   old    #8            21-3/4"                 0.0020
>   2 pd   #605-1/2      13"                     0.0005
>   CDN SW #5            12"                     0.0008
 
    I changed the support points in like that from 24", adjusted
    the loading and measuring points in to where you put them, all
    only for my Stanley #8, and got a deflection of 0.007 inches.
 
    Thanks for measuring those things, this is gonna get more
    interesting as I keep working on it ( in my spare moments when
    I'm exhausted for anything else. )  Yep, there's gonna be more
    parts...  this stuff is too accessible to be the least bit
    controversial.
 
    My estimates were geared to be maximal estimates, i.e. I tried
    to give the people who say planes flex like crazy as much benefit
    of a doubt as possible.  I.e., I'd feel safe saying that you
    wouldn't see any _more_ deflection than that, pretty much.  To
    that end, I had set the side height to be roughly it's minimal
    average value of an inch.  The side height of mine peaks at 
    2 3/8 inches, and is above 1.5 inches for a good 5 inches of
    plane length.  If you set the "effective side height" to 1.5
    inches, my earlier calculation matches your result of 0.002 inches
    exactly.  I think a good estimate of the effective side height is
    one of the most critical things here, as far as that calculation
    went, barring a more elaborate analysis.
 
    Stay tuned for Parts III+, and notes on the earlier replies to
    this thread.
 
    Doug Dawson
    dawson@p...
 
 

[tamp]

It's good to see the members of the flat-sole society having their tea out
on the front porch today (not that I believe any of this stuff) :^).

Paddy GM/ENB/Owns no optically-ground planes.

After Jeff's reply, what else is there to say except that the theoretical
deflection calculations wouldn't apply to  the tool in use, ie with
something under it!

A simple oversight I guess. Maybe the galoot equivalent of Schrodinger's cat. 

Eric

 
> After Jeff's reply, what else is there to say except that the theoretical
> deflection calculations wouldn't apply to  the tool in use, ie with
> something under it!

   It's a multi-step thing, this plane-performance stuff.  There are
   a number of different things to consider.  So far we've looked at
   cast-iron flexibility, the effect of planing pressure on the 
   flatness of the worksurface, and we've seen a simple estimate of
   what the plane-sole flatness requirements would be if all there
   was to it was geometry.  But there are a number of other issues
   to consider as well.  None of these things can be taken in 
   isolation as saying anything one way or the other.  It's about
   understanding what's going on physically when you plane.  One can
   be forgiven for being curious about that, of course.  The things
   that have shown up in this thread so far are just preliminary,
   nothing is definitive there.  It's the system as a whole that
   rules the outcome.

   BTW, there's some of this stuff that every woodworker should
   know, regardless of whether they're even into handtools or not.
   The behaviour of boards under load, etc..  It's useful.

   Doug Dawson
   dawson@p...

 
   Earlier Steve Turadek wrote,

> but then you point out that user-pressure won't deflect a smoother quite so
> much.  true.  and also we're pretty much disregarding the fact the sole may
> be *twisted* instead of simply concave, front-to-back.

   I'm assuming no twist.  Thinking in terms of the symmetry of the
   plane body, the left and right sides are gonna deform after 
   manufacture in a similar manner, unless there's something unusual
   about the metal, or the plane body was cast assymetrically.  So,
   while twist may be there, it's gonna be just a small fraction of
   cancavity along the plane's length, methinks typically for a plane
   of any quality.

   ( Anybody seen otherwise?  It could happen... )

   Wooden planes are a different story, though.

> I think the way the
> plane is constructed, you've got a heavy frog tending to resist any user
> pressure in the short axis of the plane (but also trying to flatten it in
> that axis as well.)  this is all getting complicated.

   You have to break the problem down into small pieces, then it's 
   not so hard.  ... For the #8 I'm ignoring the frog, because it's
   only a tenth the length of the plane.
 
> back to the original example.  okay, you've got an gross concave surface
> and user pressure holds the center flat to the work.  but how much pressure
> is exerted against the work *at the mouth* after this deflection is
> flattened out?  in the example given: 0.  in effect, the mouth has *no*
> effect.

   Any pressure _above_ the straightening pressure will apply to
   the mouth.

   BTW, in case you haven't yet guessed what's next... :  Wood deforms
   as well.  As you apply pressure to the ( assumed concave ) plane,
   the areas at the toe and heel in contact with the wood will compress
   the wood downwards, thus lowering the overall height of the mouth
   above the wood, and reducing any amount of flexure required to
   maintain mouth contact.

> anyway, personally, I don't have any planes this far out of tune, so I
> can't speak to how well they do or don't perform.  I have had planes with
> unflat bottoms and I've tuned (flattened) them and they seem to work
> better.  I haven't done this with a #8, but I've done if with a couple of
> modern blocks and have found them vastly improved.

   Ah, the block plane!  Now that's a situation where flexure will
   approach being negligable, and things like compression of the wood
   at either end of the plane ( to approach mouth contact ) will start
   to take over.

   Doug Dawson
   dawson@p...

   Just say, I'm not a flat-soler, really I'm not! etc.

   Earlier Gene wrote,
 
> Richard W wrote:
  
> > Q1)        Can I take it that, when using a No 8, I can bend it by 3 - 4
> >            thou?  What happens when, at the final stroke, I let the plane
> >            caress the surface, doing my best to apply only forwards
> >            pressure.?

> A1) No. I'd be amazed if you have the weight to hold a #8 in the normal user
> way and produce 10# direct downward force at the mouth, while managing to
> move it in any other direction.  I can't.  Also, my number came from a setup
> that would simulate a sole so concave that only the extreme toe and heel are
> in contact with wood.  Not a real life situation.

   I seem to vaguely recall an article in American Woodworker ( was it? )
   in which they'd taken a variety of Jack planes and looked at their
   sole flatness on a surface plate, and shown photographs of the result.
   But I can't for the life of me find this issue, my library is such an
   amazing rat's-nest.  Does anyone have the exact reference?  Anyway, it
   was interesting to look at, as I recall, and is applicable here.

> A1.1) >What happens when, at the final stroke, I let the plane
> > caress the surface, doing my best to apply only forwards pressure.?

   Provided you've planed the _entire_ _board_ with this same plane,
   by the time you're eventually done and are taking the final stroke,
   the board will have been shaped to conform to the radius of the
   sole, _provided_ you were always planing in the same exact direction.
   Alas, if you switched planes you'd have to start all over again,
   over the whole surface of the board, to have this turn out okay
   reliably.  One doesn't necessarily want to have to go through that.

   Doug Dawson
   dawson@p...

   Just say, I'm too lazy to not care!, etc.

   Earlier Jeff wrote,

> Doug Dawson wrote:

> ~  It's been a slow day on the oldtools list, so I had to fill out
> ~  my dinner break with _something_... :-)
> ~
> ~  So far we have a concrete estimate by the ISO of acceptable standards
> ~  for plane sole flatness.  As far as I know, the derivation of that
> ~  standard did not take into account sole flexure, i.e. bending of the
> ~  sole of the plane under downward pressure applied to the plane,
> ~  through the tote, while planing.

> Now this is something new! Standards as estimates.

   It is one point of view, a way of looking at a plane sole, that
   someone might take whose mind wished to come flatly to rest after
   just a little ways. :-)

> ~  It's not hard to get a reasonable idea of what effect that would
> ~  have, for a cast iron bench plane, using standard structural
> ~  analysis foo.  The exact analysis allowing for precise shape and
> ~  so forth is tedious, but we can get a reasonable ballpark figure.
>
> What is a "Ballpark" please?

   A ballpark estimate is an estimate of something, the estimated
   value of which is "in the same ballpark" as the actual number.
   An allusion to Our National Pastime, hitting balls in a baseball
   park.  But as Ollie North said, Nicaraguans don't play baseball...
   Just say, not wildly off.

   And note again, I was estimating the _maximal_ deflection you'd
   likely ever see.

   [ tamp, and tamp of my original post on this... ]

> Messing about, plane in hand, with bathroom scales on the bench top
> (to examine another theory), I found that with my puny 11 stone (154
> pounds) I can register about 90 lbs on the scale. According to Doug's
> figures, this makes a deflection of about 180thou or 0.18 inches. Cor
> blimey! Cor luv a duck!

   I'd made a typo in that _original_ formula, which I corrected to
   the list later that day.  That figure should actually be a quarter
   of what I originally wrote, or 1/32" in your example, again as an
   estimate of what you'd see no more deflection than.  If you tried
   the experiment of doing this, i.e. the setup as I described, I
   expect you'd likely see a deflection which would be a significant
   fraction of that.

   [ tamp ]                                                             
 
> Of course we should pay tribute to this static analysis. What I wonder
> would be the outcome of an analysis of the dynamic situation, taking
> into account of the need progressively to adjust (no split infinitives
> here, no siree) for temperature rises. To what extent does suction on
> non-corrugated soles affect this deflection, one gently enquires?

   I was wondering if you could describe what you mean by static versus
   dynamic in your useage.  I know what they mean, but I want to make
   sure we're talking about the same thing.

   ...  Temperature rises: possibly a factor, to the extent that you
   had differential heating, i.e. one part of the plane body a
   significantly different temperature than the other, which might
   result in some warpage you might or might not need to take into
   account.  But cast iron is a _reasonably_ good conductor of heat,
   so a large component of the effect of heat would be uniform over
   the plane body, and so an unlikely culprit in any deformation.

   ...  Corrugated soles: not so much of an effect, on raw deflection,
   given that the resistance to deflection mainly comes from the sides
   of the plane.  Suction?  We could estimate it to get it out of
   the way...  I'm assuming a smooth-bottomed plane at the moment
   for the sake of concreteness.
 
> However, from where does the ISO's Chief Savant derive this concept of
> /natural/ concavity (undefined in extent) of a plane sole? Observation
> and testing of a statistically significant quantity? Workshop lore?
> Exchange of data at one of these swapmeets one reads about? Mind you,
> most of us from time to time can get preoccupied with certain
> concavities, and convexities for that matter, especially when
> well-dined on a Sunday evening.

   Assuming that a plane sole is not twisted, which in a properly
   constructed plane would be a secondary effect next to curvature
   along the length of the plane, you have three things that could
   happen:  1) The plane sole could be perfectly flat.  This case
   is moot.  2) The location of the mouth ( vertically ) could be
   _below_ the plane occupied by the heel and toe.  This is convexity,
   which we're not considering at the moment, but may later.
   3) The location of the mouth could be _above_ the plane occupied
   by the heel and toe.  This is the situation of concavity that
   we're currently discussing.  ...These three cases are all there is.

> Perhaps next time Doug is well-dined, he might favour us with his
> formula for natural concavity, no doubt including factors such as age,
> (the plane not the owner!) nature, location and area of patent marks,
> length, presence of a complete decal (whatever that is), whether it
> has a low or high knob, factors for rosewood and otherwise, japanning
> percentage, whether bedrock or not and so on. Minor factors such as
> mis-match of knob and tote patterns can perhaps indulgently be
> excused.

   I'm assuming that you have a plane that is concave.  I don't give
   a dog's breakfast how it got that way...  That's not presently an
   issue here.

   When are people think about this problem, it's possible for them
   to get sidetracked by misjudging the importance of some things
   compared to others, such as the location of a decal, etc..  You
   have to be smart about this, and try to look for the most
   significant elements.  You can quickly get a reasonable idea of
   the effects of various things, to within an order of magnitude,
   and either decide whether they matter, or move on to some other
   factor that matters more.  Don't let yourself get too intimidated.

   Doug Dawson
   dawson@p...

   Just say, Jeff, you're one of the people who originally suggested
   that all this would be a good thing to look into, etc.

Doug writes (Taken completely out of context.  Can't say I
understood the rest.  It's pretty late.) :

>   When are people think about this problem, it's possible for them
>   to get sidetracked by misjudging the importance of some things
>   compared to others, such as the location of a decal, etc.. 

I'd wager that for some around here, the location of a decal is 
quite important ! :-)  Like my hand grinder.  It has a pretty neat
decal near the bottom (I think it's in the right place) portraying
a big gear with a couple of smaller ones engaged.  Unfortunately 
it is torn a bit in the middle.

Paul     

Montreal (Quebec)