Browse Groups

• ## Re: [multimachine] Dave, a question

(9)
• NextPrevious
• ... Okay, let s start with three plates, called A , B , and C . Put A on a pretty flat surface, cover it with an abrasive slurry and move it in a figure
Message 1 of 9 , Nov 30, 2010
View Source
On 11/30/2010 02:35 PM, Pat wrote:
> Hi Dave,
> In the past you have mentioned the 3 plate method for grinding flat surfaces. Could this same technique be used to finish the sides of 2 pieces of hot rolled plate? How to do it?
>
> Pat
>
Okay, let's start with three plates, called "A", "B", and "C". Put "A"
on a pretty flat surface, cover it with an abrasive slurry and move it
in a figure eight motion. After a few minutes the two plates should
both show marks where the abrasive cuts metal. Now move "B" to the
bottom and repeat the process with "C". Repeat with "C" on the bottom
and "A" on top. Continue until all three plates show the same pattern.

What happens is that plate "A" initially goes concave and "B" goes
convex. When "B" goes on the bottom, it goes closer to flat, slightly
less concave than "A" became. After a while, all three plates will show
the same marking pattern, and will do so with very little lapping, then
the plates are flat. Now you can do the other side. The problem is
that the plates may not have parallel surfaces, but they will be VERY flat.

The abrasive grit will let you determine how flat, once you start to see
flatness on the order of molecular level and need an SEM, you have gone
too far! This technique is used to grind telescope mirrors and optical
flats to a small fraction of the wavelength of light, see
http://www.youtube.com/watch?v=Ac2lLvgPYx4 (and the other parts) for
details. By the time you get to part 3, you will finally see it work,
and it will work well!

Dave 8{)
• Hi,     for small bits of plate I have used a piece of plate glass and grinding paste .First shown this years ago for  geting a flat surface on 2stroke
Message 2 of 9 , Dec 1 2:46 AM
View Source
Hi,
for small bits of plate I have used a piece of plate glass and grinding paste .First shown this years ago for  geting a flat surface on 2stroke alloy cylinder heads .

Mikeafloat

From: David G. LeVine <dlevine@...>
To: multimachine@yahoogroups.com
Sent: Wed, 1 December, 2010 5:19:15
Subject: Re: [multimachine] Dave, a question

On 11/30/2010 02:35 PM, Pat wrote:
> Hi Dave,
> In the past you have mentioned the 3 plate method for grinding flat surfaces. Could this same technique be used to finish the sides of 2 pieces of hot rolled plate? How to do it?
>
> Pat
>
Okay, let's start with three plates, called "A", "B", and "C".  Put "A"
on a pretty flat surface, cover it with an abrasive slurry and move it
in a figure eight motion.  After a few minutes the two plates should
both show marks where the abrasive cuts metal.  Now move "B" to the
bottom and repeat the process with "C".  Repeat with "C" on the bottom
and "A" on top.  Continue until all three plates show the same pattern.

What happens is that plate "A" initially goes concave and "B" goes
convex.  When "B" goes on the bottom, it goes closer to flat, slightly
less concave than "A" became.  After a while, all three plates will show
the same marking pattern, and will do so with very little lapping, then
the plates are flat.  Now you can do the other side.  The problem is
that the plates may not have parallel surfaces, but they will be VERY flat.

The abrasive grit will let you determine how flat, once you start to see
flatness on the order of molecular level and need an SEM, you have gone
too far!  This technique is used to grind telescope mirrors and optical
flats to a small fraction of the wavelength of light, see
http://www.youtube.com/watch?v=Ac2lLvgPYx4 (and the other parts) for
details.  By the time you get to part 3, you will finally see it work,
and it will work well!

Dave  8{)

------------------------------------

-------------
We have a sister site for files and pictures dedicated to concrete machine framed machine tools. You will find a great deal of information about concrete based machines and the inventor of the concrete frame lathe, Lucian Ingraham Yeomans. Go to http://groups.yahoo.com/group/Multimachine-Concrete-Machine-Tools/

Also visit the Joseph V. Romig group for even more concrete tool construction, shop notes, stories, and wisdom from the early 20th Century.
http://tech.groups.yahoo.com/group/romig_designs/

<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/multimachine/

<*> To change settings online go to:
http://groups.yahoo.com/group/multimachine/join
(Yahoo! ID required)

<*> To change settings via email:
multimachine-digest@yahoogroups.com
multimachine-fullfeatured@yahoogroups.com

<*> To unsubscribe from this group, send an email to:
multimachine-unsubscribe@yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/

• Hello Dave  I am understanding that Pat is asking ,how to finish the edge of the plate.  Looking at the Hot rolled flatbars I have, the edges are rounded and
Message 1 of 9 , Dec 1 5:30 AM
View Source
 Hello Dave  I am understanding that Pat is asking ,how to finish the edge of the plate.  Looking at the Hot rolled flatbars I have, the edges are rounded and not square.    KeithDeep Run PortageBack Shop" The Lizard Works" --- On Tue, 11/30/10, David G. LeVine wrote: From: David G. LeVine Subject: Re: [multimachine] Dave, a questionTo: multimachine@yahoogroups.comDate: Tuesday, November 30, 2010, 11:19 PM   On 11/30/2010 02:35 PM, Pat wrote:> Hi Dave,> In the past you have mentioned the 3 plate method for grinding flat surfaces. Could this same technique be used to finish the sides of 2 pieces of hot rolled plate? How to do it?>> Pat> Okay, let's start with three plates, called "A", "B", and "C". Put "A" on a pretty flat surface, cover it with an abrasive slurry and move it in a figure eight motion. After a few minutes the two plates should both show marks where the abrasive cuts metal. Now move "B" to the bottom and repeat the process with "C". Repeat with "C" on the bottom and "A" on top. Continue until all three plates show the same pattern.What happens is that plate "A" initially goes concave and "B" goes convex. When "B" goes on the bottom, it goes closer to flat, slightly less concave than "A" became. After a while, all three plates will show the same marking pattern, and will do so with very little lapping, then the plates are flat. Now you can do the other side. The problem is that the plates may not have parallel surfaces, but they will be VERY flat.The abrasive grit will let you determine how flat, once you start to see flatness on the order of molecular level and need an SEM, you have gone too far! This technique is used to grind telescope mirrors and optical flats to a small fraction of the wavelength of light, see http://www.youtube.com/watch?v=Ac2lLvgPYx4 (and the other parts) for details. By the time you get to part 3, you will finally see it work, and it will work well!Dave 8{)

• Thanks Keith Exactly right! I did not explain it at all well. This is especially important to me since I think that it is the last barrier to a \$250 machine
Message 1 of 9 , Dec 1 7:18 AM
View Source
Thanks Keith

Exactly right! I did not explain it at all well.
This is especially important to me since I think that it is the last barrier to a \$250 machine that can make almost all of the Mazak piece. Speaking of which, I lost the Mazak URL again! I hope someone will post it and also put it in "links".

Pat
--- In multimachine@yahoogroups.com, keith gutshall <drpshops@...>wrote:
>
> Hello Dave
> Â I am understanding that Pat is asking ,how to finish the edge of the plate.
> Â Looking at the Hot rolled flatbars I have, the edges are rounded and not square.
> Â
> Â Keith
>
> Deep Run Portage
> Back Shop
> " The Lizard Works"
>
> --- On Tue, 11/30/10, David G. LeVine <dlevine@...> wrote:
>
>
> From: David G. LeVine <dlevine@...>
> Subject: Re: [multimachine] Dave, a question
> To: multimachine@yahoogroups.com
> Date: Tuesday, November 30, 2010, 11:19 PM
>
>
> Â
>
>
>
> On 11/30/2010 02:35 PM, Pat wrote:
> > Hi Dave,
> > In the past you have mentioned the 3 plate method for grinding flat surfaces. Could this same technique be used to finish the sides of 2 pieces of hot rolled plate? How to do it?
> >
> > Pat
> >
> Okay, let's start with three plates, called "A", "B", and "C". Put "A"
> on a pretty flat surface, cover it with an abrasive slurry and move it
> in a figure eight motion. After a few minutes the two plates should
> both show marks where the abrasive cuts metal. Now move "B" to the
> bottom and repeat the process with "C". Repeat with "C" on the bottom
> and "A" on top. Continue until all three plates show the same pattern.
>
> What happens is that plate "A" initially goes concave and "B" goes
> convex. When "B" goes on the bottom, it goes closer to flat, slightly
> less concave than "A" became. After a while, all three plates will show
> the same marking pattern, and will do so with very little lapping, then
> the plates are flat. Now you can do the other side. The problem is
> that the plates may not have parallel surfaces, but they will be VERY flat.
>
> The abrasive grit will let you determine how flat, once you start to see
> flatness on the order of molecular level and need an SEM, you have gone
> too far! This technique is used to grind telescope mirrors and optical
> flats to a small fraction of the wavelength of light, see
> http://www.youtube.com/watch?v=Ac2lLvgPYx4 (and the other parts) for
> details. By the time you get to part 3, you will finally see it work,
> and it will work well!
>
> Dave 8{)
>
• Pat, I think you are already signed-up over at CNCZone.com. Take a look at this discussion:
Message 1 of 9 , Dec 1 2:40 PM
View Source
Pat,

I think you are already signed-up over at CNCZone.com. Take a look at this
discussion:

They are making V ways on an angle iron edge. I see no reason the same
setup could not be used to make a flat edge on a plate. Once roughed with
a jig and grinder, you could then use the A, B, C method to make them more
flat.

Regards,

Shannon DeWolfe
at work

> Thanks Keith
>
> Exactly right! I did not explain it at all well.
> This is especially important to me since I think that it is the last
> barrier to a \$250 machine that can make almost all of the Mazak piece.
> Speaking of which, I lost the Mazak URL again! I hope someone will post it
> and also put it in "links".
>
> Pat
> --- In multimachine@yahoogroups.com, keith gutshall <drpshops@...>wrote:
>>
>> Hello Dave
>> Â I am understanding that Pat is asking ,how to finish the edge of the
>> plate.
>> Â Looking at the Hot rolled flatbars I have, the edges are rounded and
>> not square.
>> Â
>> Â Keith
• For the sides of the two plates, that means you have 4 surfaces. If they are A,B and C,D then you could rub A+C and A+D but not C+D etc. I have read similar
Message 1 of 9 , Dec 1 10:43 PM
View Source
For the sides of the two plates, that means you have 4 surfaces. If they are A,B and C,D then you could rub A+C and A+D but not C+D etc.

I have read similar things to what Dave is describing and they always suggest 3 surfaces. What I'm not sure about is if this can be done by using the combinations of the four surfaces available here: A+C, A+D, B+C, B+D, etc. and if that would result in a flat surface... unfortunately, if my thinking is correct, then it could have the same problem as using two surfaces - the result may be concave/convex.

It is probably best to do this with 3 surfaces. Once complete, you can use those surfaces as precision references to generate more flat surfaces by using a marking compound and a scraper or file. Put a thin layer of marking on the reference and rub it on the work. This will mark the work where there are high spots. Remove these high spots with a file or scraper. Continue to repeat this until there is a consistent pattern on the work.

For marking, prussian blue oil paint works, and it can be thinned with oil. One old-world compound was 'lamp black' mixed with oil. Anything that can be spread in a thin layer and will leave a distinctive mark. The professional stuff is called 'marking blue' or 'bearing blue'.

Oh and also you probably want the sides of the plates to be parallel... not sure about that one!

James.

On Wed, Dec 1, 2010 at 4:19 PM, David G. LeVine wrote:

On 11/30/2010 02:35 PM, Pat wrote:
> Hi Dave,
> In the past you have mentioned the 3 plate method for grinding flat surfaces. Could this same technique be used to finish the sides of 2 pieces of hot rolled plate? How to do it?
>
> Pat
>
Okay, let's start with three plates, called "A", "B", and "C". Put "A"
on a pretty flat surface, cover it with an abrasive slurry and move it
in a figure eight motion. After a few minutes the two plates should
both show marks where the abrasive cuts metal. Now move "B" to the
bottom and repeat the process with "C". Repeat with "C" on the bottom
and "A" on top. Continue until all three plates show the same pattern.

What happens is that plate "A" initially goes concave and "B" goes
convex. When "B" goes on the bottom, it goes closer to flat, slightly
less concave than "A" became. After a while, all three plates will show
the same marking pattern, and will do so with very little lapping, then
the plates are flat. Now you can do the other side. The problem is
that the plates may not have parallel surfaces, but they will be VERY flat.

The abrasive grit will let you determine how flat, once you start to see
flatness on the order of molecular level and need an SEM, you have gone
too far! This technique is used to grind telescope mirrors and optical
flats to a small fraction of the wavelength of light, see
http://www.youtube.com/watch?v=Ac2lLvgPYx4 (and the other parts) for
details. By the time you get to part 3, you will finally see it work,
and it will work well!

Dave 8{)

• Pat, In order to make opposite sides of a work piece parallel and flat, you must first make a surface plate. As you once said to me, A flat surface is
Message 1 of 9 , Dec 2 5:05 AM
View Source
Pat,

In order to make opposite sides of a work piece parallel and flat, you
must first make a surface plate. As you once said to me, "A flat surface
is everything. With a flat surface you can make anything."

On the piece to have two opposite edges made flat and parallel, an
assumption is made that the piece has been cut close to, but slightly
over, the desired final dimensions. Choose an edge to work flat by
lapping using the three surface technique. Compare it to your surface
plate, doing so more often as the piece comes in closer to flat. More
than likely, some scrapping will be needed to make it as flat as is
possible. When the edge is flat place it against the surface plate,
scribe a line using a height gauge to form a guide line for filing and
scrapping the opposite edge. Work the rough edge down until it too sits
flat on the surface plate. The edges are now flat and parallel.

This method was given in an 1879 book, "The Mechanician: A Treatise on
the Construction and Manipulation of Tools" written by Cameron Knight.
Mr. Knight used many more words, but basically that is the method to be
employed. He warns that thicker plate increases the work by a
disproportionate amount. He warns too that the "learner" will do well to
pay close attention when working an edge; a small bevel anywhere along
the length of the work will require the entire edge to be reduced to
eliminate the bevel.

Mr. Knight doesn't mention it but it seems to me the job of lapping the
edges flat using the three surface technique would be made much easier
by building a jig to hold the pieces vertically. Gravity would then be

For the method described, search Google books with the phrase "making
edges straight and parallel". It is the first hit. For our friends not
in the USA, the book is here too:
http://www.archive.org/details/mechaniciantreat00knigrich
but you will need to download book. The method is described beginning on
page 209. Plates and drawings are located in the back of the book.

Regards,

Mr. Shannon DeWolfe
--I've taken to using Mr. because my name misleads folks on the WWW. I am a 54 year old fat man.

On 12/2/2010 12:43 AM, James Bishop wrote:
> Oh and also you probably want the sides of the plates to be
> parallel... not sure about that one!
• ... Precisely. The trick is that if A shows no gaps to B, and B shows no gaps to C, and C shows no gaps to A, it is probably flat. The big issue is that it
Message 1 of 9 , Dec 2 10:21 AM
View Source
On 12/02/2010 01:43 AM, James Bishop wrote:
For the sides of the two plates, that means you have 4 surfaces. If they are A,B and C,D then you could rub A+C and A+D but not C+D etc.

I have read similar things to what Dave is describing and they always suggest 3 surfaces. What I'm not sure about is if this can be done by using the combinations of the four surfaces available here: A+C, A+D, B+C, B+D, etc. and if that would result in a flat surface... unfortunately, if my thinking is correct, then it could have the same problem as using two surfaces - the result may be concave/convex.

Precisely.  The trick is that if A shows no gaps to B, and B shows no gaps to C, and C shows no gaps to A, it is probably flat.  The big issue is that it may not be at right angles or parallel to anything else.  Let's assume the A:B:C relationship holds on dovetails, they are very flat, but hardly parallel.

It is probably best to do this with 3 surfaces. Once complete, you can use those surfaces as precision references to generate more flat surfaces by using a marking compound and a scraper or file. Put a thin layer of marking on the reference and rub it on the work. This will mark the work where there are high spots. Remove these high spots with a file or scraper. Continue to repeat this until there is a consistent pattern on the work.

Actually, any odd number will work, but even ones won't.  Let's assume two spherical surfaces (A:C) and two cups (B:D) with the same radius of curvature.  When comparing A to B, there is no gap, B to C no gap, C to D no gap and D to A no gap, but A to C might show a huge gap.

For marking, prussian blue oil paint works, and it can be thinned with oil. One old-world compound was 'lamp black' mixed with oil. Anything that can be spread in a thin layer and will leave a distinctive mark. The professional stuff is called 'marking blue' or 'bearing blue'.

Oh and also you probably want the sides of the plates to be parallel... not sure about that one!

James.

I am, there is no way that two sided can be guaranteed to be parallel, or even perpendicular, using this technique.

Sorry to be the party pooper here.

Dave  8{)

Your message has been successfully submitted and would be delivered to recipients shortly.
• Changes have not been saved
Press OK to abandon changes or Cancel to continue editing
• Your browser is not supported
Kindly note that Groups does not support 7.0 or earlier versions of Internet Explorer. We recommend upgrading to the latest Internet Explorer, Google Chrome, or Firefox. If you are using IE 9 or later, make sure you turn off Compatibility View.