| Lathe Bed Preparation |
| :: Kahale & Martin AP Machine :: |
A 14" bed extension kit was ordered from Chis Woods at Little Machine Shop, part # 1928. A visual inspection showed
it had survived shipment well with but a slight mar in the flat of the "V" way where the containment straps lay, a cosmetic
defect only. Seig must paint these rather rapidly as nearly all critical machined surfaces had some rather thick paint
slopped about on them and this would have to go. The worst of it was in the area inside the webbing where the tail stock
would register. A hard place to paint or to strip of paint and days would be spent in this task. Stories of large voids in
castings covered by fillers and copious amounts of paint had me already tilting toward a complete bed strip anyway so
not much excuse was required. Some areas had little or no paint and a good deal of casting sand was painted over. In
addition there were lots of edges and corners that had missed the de-burring operation.
First a critical dimensional check was needed to see if the casting even warranted this type of effort . The beds are
ground to dimension and to do so a point of reference for the grinder is needed and this would be the foot area. This
makes it fairly straight forward to qualify the casting by placing it on a surface plate first checking for rock, indicating a
twist and using a height gauge checking for bow or warp and flat/level relative to the surface plate and the "V" way
relative to the flat way. The "V" requires multiple checks as angles are involved and the top of the "V" in truncated. This
was done with the top half of a bed lock as shown in the photo below. Once the "V" is qualified then direct micrometer
measurement from the top of the "V" to the lower bearing area for the saddle gib area is assured. If the "V" is not
qualified, as in an existing machine, then the bed lock block and micrometer is required for thickness checks.
it had survived shipment well with but a slight mar in the flat of the "V" way where the containment straps lay, a cosmetic
defect only. Seig must paint these rather rapidly as nearly all critical machined surfaces had some rather thick paint
slopped about on them and this would have to go. The worst of it was in the area inside the webbing where the tail stock
would register. A hard place to paint or to strip of paint and days would be spent in this task. Stories of large voids in
castings covered by fillers and copious amounts of paint had me already tilting toward a complete bed strip anyway so
not much excuse was required. Some areas had little or no paint and a good deal of casting sand was painted over. In
addition there were lots of edges and corners that had missed the de-burring operation.
First a critical dimensional check was needed to see if the casting even warranted this type of effort . The beds are
ground to dimension and to do so a point of reference for the grinder is needed and this would be the foot area. This
makes it fairly straight forward to qualify the casting by placing it on a surface plate first checking for rock, indicating a
twist and using a height gauge checking for bow or warp and flat/level relative to the surface plate and the "V" way
relative to the flat way. The "V" requires multiple checks as angles are involved and the top of the "V" in truncated. This
was done with the top half of a bed lock as shown in the photo below. Once the "V" is qualified then direct micrometer
measurement from the top of the "V" to the lower bearing area for the saddle gib area is assured. If the "V" is not
qualified, as in an existing machine, then the bed lock block and micrometer is required for thickness checks.
Pictures are worth a thousand words! As you can see the project is several steps ahead of the site. This bed checked
out dead flat and all relative positions in alignment except that the lower bearing block area for the saddle gibs, which
are milled and not ground, displayed slight thickness variations which will be corrected latter when the saddle is fit to the
bed.
After the checks were done the entire bed was stripped inside and out. Machined exterior surfaces were masked with
Duct Tape to prevent nicks and dings while the rough handling was going on, there is no gentile way to do this. A spray
type stripper for epoxies was used as gel types were quite ineffective. Small brass bristle brushes were required in the
rough areas of the internal webbing and motor mount areas and repeated coats required even then. A good washing in
a bucket of mineral spirits followed as well as allot of fine filing of the burrs and sharp edges.
No serious voids were uncovered but a good deal of casting sand was eliminated during this step. The part, sand cast,
uses about a dozen cores, none of which line up well. Ergo there is a good deal of very soft filler used in the worst areas.
As you can see from the photo just to the left of the height gauge most surfaces not ground or left natural, are fly milled
and not gently. None of the red you see in the photo was present after stripping. This is a thin coat of auto body glaze
after sanding with a block sander. The difference in hue is caused by oils used while fitting the saddle and will have to
be cleaned latter before painting. The interior of the webbing and motor mount areas are already painted with Glyptal
electric motor insulating paint. It's tough and quite resistant to oils.
out dead flat and all relative positions in alignment except that the lower bearing block area for the saddle gibs, which
are milled and not ground, displayed slight thickness variations which will be corrected latter when the saddle is fit to the
bed.
After the checks were done the entire bed was stripped inside and out. Machined exterior surfaces were masked with
Duct Tape to prevent nicks and dings while the rough handling was going on, there is no gentile way to do this. A spray
type stripper for epoxies was used as gel types were quite ineffective. Small brass bristle brushes were required in the
rough areas of the internal webbing and motor mount areas and repeated coats required even then. A good washing in
a bucket of mineral spirits followed as well as allot of fine filing of the burrs and sharp edges.
No serious voids were uncovered but a good deal of casting sand was eliminated during this step. The part, sand cast,
uses about a dozen cores, none of which line up well. Ergo there is a good deal of very soft filler used in the worst areas.
As you can see from the photo just to the left of the height gauge most surfaces not ground or left natural, are fly milled
and not gently. None of the red you see in the photo was present after stripping. This is a thin coat of auto body glaze
after sanding with a block sander. The difference in hue is caused by oils used while fitting the saddle and will have to
be cleaned latter before painting. The interior of the webbing and motor mount areas are already painted with Glyptal
electric motor insulating paint. It's tough and quite resistant to oils.
You can see in the picture to the right above, at the far right a shiny area. This is ground flat to the foot area and not
painted to avoid elevation changes latter when mounted. Look into the web area of this same photo at the register for
the tail stock hold down, this area was covered in paint from the factory. The left photo showing paint work in the
motor mount area. These areas were painted first, right after the mineral wash as they are very rough and cleaning
oils and grit latter would present problems and prevent good coatings adhesion. Attention to the details. The Duct
Tape will remain in place until we're ready to mount it to the bench plate to protect it.
Painting the interior is a very tedious task and very time consuming. More time is required than the "open time" of the
paint which is in a xylene solvent so just a touch of mineral spirits was added to slow the dry rate and prevent
ghosting of edges. Masking the machined areas is nearly impossible so a steady hand, small art brushes, lots of
angles of attack are the weapons to prevent re-coating the areas you just spent so much time stripping. I actually
over did the solvent cut and it took nearly two weeks for this 24 hour cure paint to dry. This could have been forced
dried in an oven (if I could have convinced the wife the smell in her kitchen would disappear, really!)
The bed as delivered weighed in at 25 lbs. After stripping and knocking sand out of it and removing soft fillers, 24
lbs. The 14" beds have several more internal webs and the right foot is fully cast unlike my short bed machine. The
motor mount area is wider to accommodate the larger motors normally fit to these machines as well. Both feet of the
casting are also longer.
With the bed initial preparation completed and measurements qualified the next task is to fit the saddle assembly.
This was a job and deserves its own page.
Don't under estimate the importance of all the detail work. Everything references from the bed of the machine, if it
isn't right, nothing else will be either. T'is the very foundation of your project.
painted to avoid elevation changes latter when mounted. Look into the web area of this same photo at the register for
the tail stock hold down, this area was covered in paint from the factory. The left photo showing paint work in the
motor mount area. These areas were painted first, right after the mineral wash as they are very rough and cleaning
oils and grit latter would present problems and prevent good coatings adhesion. Attention to the details. The Duct
Tape will remain in place until we're ready to mount it to the bench plate to protect it.
Painting the interior is a very tedious task and very time consuming. More time is required than the "open time" of the
paint which is in a xylene solvent so just a touch of mineral spirits was added to slow the dry rate and prevent
ghosting of edges. Masking the machined areas is nearly impossible so a steady hand, small art brushes, lots of
angles of attack are the weapons to prevent re-coating the areas you just spent so much time stripping. I actually
over did the solvent cut and it took nearly two weeks for this 24 hour cure paint to dry. This could have been forced
dried in an oven (if I could have convinced the wife the smell in her kitchen would disappear, really!)
The bed as delivered weighed in at 25 lbs. After stripping and knocking sand out of it and removing soft fillers, 24
lbs. The 14" beds have several more internal webs and the right foot is fully cast unlike my short bed machine. The
motor mount area is wider to accommodate the larger motors normally fit to these machines as well. Both feet of the
casting are also longer.
With the bed initial preparation completed and measurements qualified the next task is to fit the saddle assembly.
This was a job and deserves its own page.
Don't under estimate the importance of all the detail work. Everything references from the bed of the machine, if it
isn't right, nothing else will be either. T'is the very foundation of your project.
 | |  |
| Fitting the Saddle |
Using the height gauge a near side reference is marked to locate it in the same position each time. (Picture above
to the left) Height is set from the base of the compound dove register as this is referenced not only to the bed but will
be the reference for the compound later when fitted. Of course cleanliness is next to godliness when making such
measurements. Note I used the gauge foot extension to do this to get as much distance between points as possible.
Once zero is set the height gauge is rotated to the back as in the picture on the right. A dial indicator on magnetic base
set and zeroed. Now pressure is placed upon the near side of the saddle (operator side) using the "V" as the fulcrum
which will lift the back side until it contacts the height gauge and down hill distance can now be read directly from the
dial indicator. When this was first done the back of the saddle was .018" lower than the front!
The height gauge can be moved to the opposite side to detect difference in compound dove base height. In this case
a few thou in difference.
Next the measurement equipment is removed form the bed and finger pressure applied to opposite corners, such as
front left and right rear to detect any tendency to rock and there was. Use of the dial indicator or a feeler gauge can
quantify the amount. Roughly .005" in this case.
Think about this for a minute. It is possible to do nothing and adjust the saddle gibs to get minimum movement and
free action and be fooled into thinking you have a well fit saddle. In so doing there will be single points of contact on
both the back flat and the near "V". Three things will happen when the machine is placed under cutting loads.
Next one must identify the high spots and points of contact. Layout die applied to the saddle "V" and back flat and then
setting it on the bed and sliding it back a forth a few inches will mark those places.
As side to side was pretty close I assumed the "V" to be pretty close in that plane if but considerably high and the
major problem area to be the back flat. As it turned out the "V" area was making contact on both sides but with little
area in contact. The back flat however removed next to nothing of the bluing. Just a light pencil lead width concentrated
on the high corner. As it was so far out I used a medium sized second cut file to knock it down. Easy, a few licks at a
time and checking. This whole process is slow and painful. Once the rocking motion was close to nil a piece of wet/dry
paper, 80 grit was placed grit side up on the bed under the flat. A piece of paper over the "V" to protect it and the saddle
worked over this until all trace of rock was removed. Of course this makes the down hill condition even worse but full
line contact made across the flat which also corrected the side to side difference for compound dove register.
to the left) Height is set from the base of the compound dove register as this is referenced not only to the bed but will
be the reference for the compound later when fitted. Of course cleanliness is next to godliness when making such
measurements. Note I used the gauge foot extension to do this to get as much distance between points as possible.
Once zero is set the height gauge is rotated to the back as in the picture on the right. A dial indicator on magnetic base
set and zeroed. Now pressure is placed upon the near side of the saddle (operator side) using the "V" as the fulcrum
which will lift the back side until it contacts the height gauge and down hill distance can now be read directly from the
dial indicator. When this was first done the back of the saddle was .018" lower than the front!
The height gauge can be moved to the opposite side to detect difference in compound dove base height. In this case
a few thou in difference.
Next the measurement equipment is removed form the bed and finger pressure applied to opposite corners, such as
front left and right rear to detect any tendency to rock and there was. Use of the dial indicator or a feeler gauge can
quantify the amount. Roughly .005" in this case.
Think about this for a minute. It is possible to do nothing and adjust the saddle gibs to get minimum movement and
free action and be fooled into thinking you have a well fit saddle. In so doing there will be single points of contact on
both the back flat and the near "V". Three things will happen when the machine is placed under cutting loads.
- It will chatter like a jackhammer.
- It will be hard to keep in adjustment as single point contact accelerates wear.
- A tool set to center by a facing cut will high on the OD and the larger the diameter the higher it will be, again
either causing chatter or tool rub on tools with little relief.
Next one must identify the high spots and points of contact. Layout die applied to the saddle "V" and back flat and then
setting it on the bed and sliding it back a forth a few inches will mark those places.
As side to side was pretty close I assumed the "V" to be pretty close in that plane if but considerably high and the
major problem area to be the back flat. As it turned out the "V" area was making contact on both sides but with little
area in contact. The back flat however removed next to nothing of the bluing. Just a light pencil lead width concentrated
on the high corner. As it was so far out I used a medium sized second cut file to knock it down. Easy, a few licks at a
time and checking. This whole process is slow and painful. Once the rocking motion was close to nil a piece of wet/dry
paper, 80 grit was placed grit side up on the bed under the flat. A piece of paper over the "V" to protect it and the saddle
worked over this until all trace of rock was removed. Of course this makes the down hill condition even worse but full
line contact made across the flat which also corrected the side to side difference for compound dove register.
The arrow in the left photo above shows where I had Kresser Machine run a quarter inch end mill down the roof of the
"V". The reason this is needed is that the top flat of the "V" is about .200" wide (6 mm) and the relief in the roof of the
"V" is but .157" wide (4 mm). As you attempt to remove material from this area to lower the saddle a ridge is formed
that prevents progress. Size 9 foot in a size 8 shoe situation. In the right hand photo you will note a dark semi circle
near the front of the roof area. This is where the end mill broke through to the travel rest hold down. Not a big deal but
worth noting. We raised the roof .025". To be honest I didn't do this procedure straight away but it came as a
consequence after working this thing silly for a week with little progress. A bit more thought on this and the roof
height would remain as is as it is the width I was after. Could have done a better job of centering as well, we eye
balled it as he was near closing time on a Friday and doing it for free.
You can't see it in the left picture but as it is corrected now but if you could have you would have seen a space on the
back flat you could slide a match book cover under between saddle and bed.
Okay so now we have the rock out, side to side squared up and some room in the "V" to work with, now the hard part.
"V". The reason this is needed is that the top flat of the "V" is about .200" wide (6 mm) and the relief in the roof of the
"V" is but .157" wide (4 mm). As you attempt to remove material from this area to lower the saddle a ridge is formed
that prevents progress. Size 9 foot in a size 8 shoe situation. In the right hand photo you will note a dark semi circle
near the front of the roof area. This is where the end mill broke through to the travel rest hold down. Not a big deal but
worth noting. We raised the roof .025". To be honest I didn't do this procedure straight away but it came as a
consequence after working this thing silly for a week with little progress. A bit more thought on this and the roof
height would remain as is as it is the width I was after. Could have done a better job of centering as well, we eye
balled it as he was near closing time on a Friday and doing it for free.
You can't see it in the left picture but as it is corrected now but if you could have you would have seen a space on the
back flat you could slide a match book cover under between saddle and bed.
Okay so now we have the rock out, side to side squared up and some room in the "V" to work with, now the hard part.
Wet/Dry paper is cut to width to just cover the "V" and Duct Taped in place and WD-40 applied before the saddle is set
down on it. On the flat several folds of paper are placed between the saddle and the flat. (This picture is a mock up as
the work is done now.)
The paper not only protects the un-lubricated flat but adjust the height. If you don't shim it like this or make some other
provision for height then as you sand the saddle "V" it will eventually take a new angle and not one you want. This
takes awhile to get a feel for, allot of tape and allot of sand paper and allot of paper towels allot of solvent and a lot of
patience. Work slow and measure often as was done in the beginning. How many sheets of paper? Well a sheet is
about .003: thick and were down about .021" at present plus the thickness of the sand paper so roughly eight ply will
be close and close at this point is good enough. You want to be slightly high on the side your working and work it until
your slightly low, then remove a single shim and do it again.
Measure, measure, measure...this took about an hour a night for nearly two weeks. Measurements will help you get a
feel for pressure, how many strokes to remove how much metal, what grade paper to use and so on. Nothing I can
give implicit instructions too and each case will be different.
When you get down to less than say .005" on the down hill trade the shim stock for a sheet of wet/dry on the back and
do the same on the front, trade sand paper for a paper shim. You've reduced the angle enough to get much better
contact and now you need it flat. Work both sides using a courser grade on the "V" than the flat. You'll get a feel for how
much, where and when to quit long before your in trouble. (Remember, grit side up :)
If you got all this, by now your within .0005" or so of down hill, no rocking motion, both sides within .0005" in height and
using something in the neighborhood of 320 grit paper.
Opinion varies I know but from this point I removed all paper and shim stock, wet the entire length of the bed with
WD-40 and hand applied 380 grit lapping powder working the entire length of the bed with even pressure. I'm dye
checking at this point. Clean it off and check the shadow marks it leaves on the bed. It should be even and full width
both on the "V" and on the flat, both on the bed and on the saddle. Clean it well with brake cleaner and clean white
cloth until it comes up white. The redo it with 500 grit powder. Brake cleaners do not leave an oily film behind. Repeat
the white cloth until clean.
Grits this fine remove very little material and basically just knock the tops off the asperities left behind from courser
work. It will leave a shadow line on the bed that is near impossible to get off but you can assure removal of all lapping
media by cleaning with carburettor cleaner and oiling it. Run the saddle back and forth until the oil becomes dark.
Clean and repeat until the oil and the gray is gone.
At this point it just wiggles the dial indicator so I gave both sides a few strokes with a sheet of 1500 grit, then cleaned
and again with 2000 grit, both with a liberal spray of mineral spirits and cleaned a final time.
When finished I had full contact on all contact surfaces that would remove blue (actually a Sharpie at this point) on a
single passing. When wet with WD-40 I placed the saddle at one end of the bed and gave it a tap. Don't do that, it slid
like a puck on an air hockey table crashing off the opposite end onto the bench putting a ding in the saddle that
fortunately for me was in a place of no harm. Now WD-40 isn't the normal oil to use on the bed and heavier oil does
induce more friction and the slide is less free. Acts like a joy block actually. Slides freely but near impossible to pull
straight off the bed. When I was close the dial indicator was replaced with a test indicator with four decimal resolution
and I ended with a wiggle in the indicator for down hill. Under .0002 side to side and zero rocking motion. Rechecking
with the height gauge on all surfaces and re-qualifying the "V" with the bed lock block produced no measurable
differences.
I suppose many will consider this way to smooth and flat so some words of explanation are in order.
Ever ask yourself why cars of the 30's required 40W or even 50W oil and modern Honda motors get by on 0W20?
Surface finish and precision fitting and alignment. Don't think it's power as modern engines make allot more power
per displacement that motors or yesteryear. In fact many Pro Stock motors today use 5W oils in motors near 3000 hp.
Oil needs to be thick enough to fill the asperities and no more AND provide an adequate surface film strength.
Viscosity does not measure film strength.
That said the finer and flatter the finish, the more precise the alignment (with the appropriate lubrication) the tighter
things like gibs can be set limiting motion and precluding chatter and the lighter the oil that can be used. Joy block
type finishes actually help dampen vibration using the elastic nature of the fluid as a shock absorber. Such surfaces
wear at a much lower rate due to low contact unit loading.
Now, this is what I've done and I won't preach that as gospel to anyone. IF you disagree finish to what you like and use
what lubricants you like, just get it flat and square and well aligned. Your now caught up to the current state of the
build.
3/22/2007 Rereading some of this I see I miss explanation of a very important step. Cleaning up the dimensional
differences of the bed underside gib contact area. This was done by making measurements along the way both front
and back in half inch increments writing the dimensions with felt marker on the bed. Using a 4" second cut flat file the
high areas were taken down as far as I dared. The the file wrapped in wet/dry 320 and worked until I was within
.0002", This took several days and allot of measurements. I was concerned about getting out of square but gentle
pressure working short areas at a time seemed to work just fine, even with out a guide fixture. Now the ends of the
bed on both ends taper to a smaller dimension that the area I worked but I reasoned that only the area covered by
saddle travel need be worked over. Later during gib fitting I used yet another method to finish the underside to an
immeasurable tolerance.
down on it. On the flat several folds of paper are placed between the saddle and the flat. (This picture is a mock up as
the work is done now.)
The paper not only protects the un-lubricated flat but adjust the height. If you don't shim it like this or make some other
provision for height then as you sand the saddle "V" it will eventually take a new angle and not one you want. This
takes awhile to get a feel for, allot of tape and allot of sand paper and allot of paper towels allot of solvent and a lot of
patience. Work slow and measure often as was done in the beginning. How many sheets of paper? Well a sheet is
about .003: thick and were down about .021" at present plus the thickness of the sand paper so roughly eight ply will
be close and close at this point is good enough. You want to be slightly high on the side your working and work it until
your slightly low, then remove a single shim and do it again.
Measure, measure, measure...this took about an hour a night for nearly two weeks. Measurements will help you get a
feel for pressure, how many strokes to remove how much metal, what grade paper to use and so on. Nothing I can
give implicit instructions too and each case will be different.
When you get down to less than say .005" on the down hill trade the shim stock for a sheet of wet/dry on the back and
do the same on the front, trade sand paper for a paper shim. You've reduced the angle enough to get much better
contact and now you need it flat. Work both sides using a courser grade on the "V" than the flat. You'll get a feel for how
much, where and when to quit long before your in trouble. (Remember, grit side up :)
If you got all this, by now your within .0005" or so of down hill, no rocking motion, both sides within .0005" in height and
using something in the neighborhood of 320 grit paper.
Opinion varies I know but from this point I removed all paper and shim stock, wet the entire length of the bed with
WD-40 and hand applied 380 grit lapping powder working the entire length of the bed with even pressure. I'm dye
checking at this point. Clean it off and check the shadow marks it leaves on the bed. It should be even and full width
both on the "V" and on the flat, both on the bed and on the saddle. Clean it well with brake cleaner and clean white
cloth until it comes up white. The redo it with 500 grit powder. Brake cleaners do not leave an oily film behind. Repeat
the white cloth until clean.
Grits this fine remove very little material and basically just knock the tops off the asperities left behind from courser
work. It will leave a shadow line on the bed that is near impossible to get off but you can assure removal of all lapping
media by cleaning with carburettor cleaner and oiling it. Run the saddle back and forth until the oil becomes dark.
Clean and repeat until the oil and the gray is gone.
At this point it just wiggles the dial indicator so I gave both sides a few strokes with a sheet of 1500 grit, then cleaned
and again with 2000 grit, both with a liberal spray of mineral spirits and cleaned a final time.
When finished I had full contact on all contact surfaces that would remove blue (actually a Sharpie at this point) on a
single passing. When wet with WD-40 I placed the saddle at one end of the bed and gave it a tap. Don't do that, it slid
like a puck on an air hockey table crashing off the opposite end onto the bench putting a ding in the saddle that
fortunately for me was in a place of no harm. Now WD-40 isn't the normal oil to use on the bed and heavier oil does
induce more friction and the slide is less free. Acts like a joy block actually. Slides freely but near impossible to pull
straight off the bed. When I was close the dial indicator was replaced with a test indicator with four decimal resolution
and I ended with a wiggle in the indicator for down hill. Under .0002 side to side and zero rocking motion. Rechecking
with the height gauge on all surfaces and re-qualifying the "V" with the bed lock block produced no measurable
differences.
I suppose many will consider this way to smooth and flat so some words of explanation are in order.
Ever ask yourself why cars of the 30's required 40W or even 50W oil and modern Honda motors get by on 0W20?
Surface finish and precision fitting and alignment. Don't think it's power as modern engines make allot more power
per displacement that motors or yesteryear. In fact many Pro Stock motors today use 5W oils in motors near 3000 hp.
Oil needs to be thick enough to fill the asperities and no more AND provide an adequate surface film strength.
Viscosity does not measure film strength.
That said the finer and flatter the finish, the more precise the alignment (with the appropriate lubrication) the tighter
things like gibs can be set limiting motion and precluding chatter and the lighter the oil that can be used. Joy block
type finishes actually help dampen vibration using the elastic nature of the fluid as a shock absorber. Such surfaces
wear at a much lower rate due to low contact unit loading.
Now, this is what I've done and I won't preach that as gospel to anyone. IF you disagree finish to what you like and use
what lubricants you like, just get it flat and square and well aligned. Your now caught up to the current state of the
build.
3/22/2007 Rereading some of this I see I miss explanation of a very important step. Cleaning up the dimensional
differences of the bed underside gib contact area. This was done by making measurements along the way both front
and back in half inch increments writing the dimensions with felt marker on the bed. Using a 4" second cut flat file the
high areas were taken down as far as I dared. The the file wrapped in wet/dry 320 and worked until I was within
.0002", This took several days and allot of measurements. I was concerned about getting out of square but gentle
pressure working short areas at a time seemed to work just fine, even with out a guide fixture. Now the ends of the
bed on both ends taper to a smaller dimension that the area I worked but I reasoned that only the area covered by
saddle travel need be worked over. Later during gib fitting I used yet another method to finish the underside to an
immeasurable tolerance.
| Fitting the Saddle Gibs |
Factory Gibs leave a bit to be desired. Hard to adjust and just too soft and limp to be effective. This is obvious if you
use shims in lieu of the dog point screws for adjustment. My first attempt at adjustment, with the screws, bent the gib
enough to give but point contact on the two far corners. Second attempt using half hard 260 brass shim stock dented
the gib strips so a new plan need be devised.
After careful measurement of the existing gibs and consulting others drawing I determined that standard .750" X .250"
precision ground A-2 tool steel would get the correct envelope. This material comes in 18" lengths and as I have two
machines this leaves me with enough material in surplus for the kerfs. As purchased it comes ground to tolerance
and about 20 to 25 Rockwell C making it machinable. As the notes in the drawing are difficult to read these are being
made to drawing, then heat treated and tempered to 55 to 62 Rockwell C.
Heat treat normally distorts the material. A-2 is less susceptible to this and the reason the material was chosen.
Hardened parts will then be surface ground to be no more than .0002" from parallel and ditto for flat. They will then be
treated to a flash (.0003") of Industrial Hard Chrome. This surface finish does several things. First it is extremely wear
resistant. Secondly provides corrosion resistance and lastly has a coefficient of friction near that of Teflon. Drawings,
material list and construction drawings were submitted 02/22/2007 to Kresser Machine with a lead time of about 10
days. Kresser will do the machine and grinding work and the heat treat and chroming will be farmed out to third
parties.
Unfortunately TC88, the program used for this drawing does not export to a JPEG so I had to scan it, format it and
process it which is the reason for the slight distortion, sorry. All dimensions are legible though thus useful.
Factory gibs have a potential contact area of about .985" square per side. As fit they actually delivered about .0625"
squared with the point contact from all the bending. This set being made will have a useful 1.2" squared contact area
or a bit better than a 20% increase over factory plan and 190% more than as she sits. Unlike the factory gummy bear
gibs, these will be stiff enough to not flex under load, period. The friction component will be about half. Wear rates
should be but a fraction thus I have little concern about the nature of difficulty with the shim style adjustment. Hence,
taper gibs are now off the table.
use shims in lieu of the dog point screws for adjustment. My first attempt at adjustment, with the screws, bent the gib
enough to give but point contact on the two far corners. Second attempt using half hard 260 brass shim stock dented
the gib strips so a new plan need be devised.
After careful measurement of the existing gibs and consulting others drawing I determined that standard .750" X .250"
precision ground A-2 tool steel would get the correct envelope. This material comes in 18" lengths and as I have two
machines this leaves me with enough material in surplus for the kerfs. As purchased it comes ground to tolerance
and about 20 to 25 Rockwell C making it machinable. As the notes in the drawing are difficult to read these are being
made to drawing, then heat treated and tempered to 55 to 62 Rockwell C.
Heat treat normally distorts the material. A-2 is less susceptible to this and the reason the material was chosen.
Hardened parts will then be surface ground to be no more than .0002" from parallel and ditto for flat. They will then be
treated to a flash (.0003") of Industrial Hard Chrome. This surface finish does several things. First it is extremely wear
resistant. Secondly provides corrosion resistance and lastly has a coefficient of friction near that of Teflon. Drawings,
material list and construction drawings were submitted 02/22/2007 to Kresser Machine with a lead time of about 10
days. Kresser will do the machine and grinding work and the heat treat and chroming will be farmed out to third
parties.
Unfortunately TC88, the program used for this drawing does not export to a JPEG so I had to scan it, format it and
process it which is the reason for the slight distortion, sorry. All dimensions are legible though thus useful.
Factory gibs have a potential contact area of about .985" square per side. As fit they actually delivered about .0625"
squared with the point contact from all the bending. This set being made will have a useful 1.2" squared contact area
or a bit better than a 20% increase over factory plan and 190% more than as she sits. Unlike the factory gummy bear
gibs, these will be stiff enough to not flex under load, period. The friction component will be about half. Wear rates
should be but a fraction thus I have little concern about the nature of difficulty with the shim style adjustment. Hence,
taper gibs are now off the table.
Use the above drawing for bolt spacing and width from bolt center to working edge only. See text for
differences from drawing to final production pieces. I'll get around to an updated drawing sooner or latter ;)
Gibs: Plan B
3/17/2007 I feel a need to rewrite this section after the work done the last few weeks. I made the gibs close to
the drawing from A-2 house scraps except thicker .350", and omitting the center hole. That made it 3.15" between
centers on the two remaining holes. The thinking on this was that if I used a shim between the cap screws and the
gib was stiff then the center hole was not required and actually in the way. All other dimensions remained in tact.
The manufacture of the gibs was very straight forward and went off without a hitch. Installing them was another
matter. I bought some brass shim stock and a set of feeler gauges to sacrifice for same. I also acquired a few
sheets of 320 grit orbital sander media with the sticky back like they use in body shops. This turned out the be a
very good idea. Nothing is square on these machines and the gib registers were no different. Use of a single shim
proved unworkable as the registers are .001" out of parallel with the bed end for end. This translates into one end
contacting the bed and the other quite loose making the gib dig into the bed. I reasoned at this point the two shims
moved outboard of the fasteners of different thicknesses might work quite nicely and it did. Once it got it pretty
close I disassembled the unit again as I had yet a few tight spots here and there. Using the gib as a template I cut
sections of the sand paper the full length and width of the gib and applied it sticky side to gib, grit to bed and
reinstalled the shims. Running the saddle up and down the bed then hits only the high spots. If you remember I
had it down to about .0002" to begin with but when you get things this close even that is too much. Both front and
back are now with in .00005 to .0001"" flat and parallel to that same dimension from the top of the bed and that is
as fine as I can measure.
High spots gone the saddle fit a tad loose again so the shim packs were thinned a bit more. Thing is that
sometimes you need a shim in some size other than standard .0005" increments. That's where the brass stock, a
lapping plate and some 2000 grit wet/dry earns its keep. Now anyone who has ever done a bit of this close work
would tell you that even bolt tension will move things a bit and the idea is to get a shim that allows enough torque
to be applied to it to prevent loosening during use and that takes a bit of trial and error, I remade shims more than
once. I might mention that clean is highly important as well as at this level of fit even dust on a shim makes or
brakes it. I use brake cleaner and lint free rags. Oil has thickness so it need be cleaned and oiled each assembly
cycle with the oil you are committed to. I'm committed to ISO 68 way oil and the supplier is unimportant but use the
same one.
After several evenings and being quite happy with it all save for a still unacceptable level of "stick" although
"slip" was about as good as anything I've ever felt, we got into a thread on the 7 X 12 site when I asked what others
were setting there gib gaps too. I was thinking that perhaps I was just too tight. Well we didn't have an answer but
being who I am I kept on plugging on other sites I haunt and found a fellow who knows his stuff. A fellow on the
Southbend site and rebuilds machines. Long story sort, I needed more contact area on the back gib and need not
worry so much about the front as tool pressures kept it in place if the rear gib is right.
Today I remade a rear gib and made it 6" long X .4555" thick and 1.030" wide. Only the length contributes to
contact area however. Extra thickness for the extra length for stiffness and width because I was too rushed to mill it
off. This time I also used all three holes and located two stand off set screws as per original except that they are
centered perfectly and in line with the three hold downs. These were made 1/4-20. Upon installation I fit the center
bolt, placed the shims as before on the outside corners for level and gap. Then fit the two outside caps and played
the bolt tension until satisfied. NOW I ran the set screws in to make firm contact and tightened the lock nuts.
Finally, as both sides of each retaining screw are now supported those can be tightened to a sufficient level to
remain in place.
The difference is night and day. Stick is gone and slip is even better. You would have to feel it to know what I
mean and I don't think a video would make the point. A test indicator on the back shows ZERO displacement of
saddle to bed and something under .0005" on the front. Placing and inclinometer on the bed and elevating until
the saddle moved gave a slip angle of 7 degrees for a frictional coefficient of .121 "stick" and maintained motion
once underway to a 5 degrees slip angle for a friction coefficient of .089. This is about a 4/5 reduction from factory
in dynamic friction. Hardened and Hard Chromed would bring this down quite a bit more.
The above paragraph may seem counter to what I said about stick being gone and a word of explanation
required. The kinetic and dynamic test of slip angles are at very low velocity and load and represent the friction first
as a result of surface finish and level of fit and secondly as a component of lubrication. In other words the shear
rate with in the oil at low velocities is a non factor. As such stick is higher than slip. When the saddle is traversed
with the machine set flat under power, in this case finger pressure, and at a rate that represents actual speeds of
operation then the shear of the oil comes into play increasing viscosity induced dynamic drag to a point higher
than that of the at rest state. Ergo in an operational sense, drag at rest is less than drag in motion. This points out
an important flaw in sense testing. The actual difference noted above in slip testing shows that the 25% reduction
in stick to slip drag is undetectable to your senses and largely academic. However the reduction in kinetic slip
angle, which was, to start, 37 degrees thus a friction coefficient of .6 is a marked and sense-able improvement. As
this value is about three times higher than normal for cast iron on mild steel even dry it only seems logical that the
remainder above this point was due to mechanical "binding" of misfit parts and rough surfaces. Flimsy gibs,
rough finish and poor flatness, angularity etc. This entire exercise has been extremely worthwhile and not a
second of the several months it's taken to get to this point a waste.
Note the shims widely spaced and not full length but they are full width to prevent rocking. The rough
measurement for shim thickness is above each but a bit of fine tuning on a sheet of 2000 grit with thinning in
increments of .0001" per fitting was required to get it right. Of course that means I had to make each one twice and
you over shoot the dimension one step the first pass. The different thickness is required as the saddle casting is
not parallel to the bed at their mounting face.
The six inch length was a vast improvement over the standard 3.9" stock and 4" X .352" thick version. The extra
length will preclude the use of the motor guard but I don't plan on the stock location for the motor anyway. The extra
width was pure laziness at first but now that it's there I see a saddle lock possibility and perhaps a thread mount for
a DTI or as it turned out a gib with four workable wear faces.
Note that the area for the cross slide screw is milled out in anticipation of a thrust bearing in the collar later. The
extra length is .250" and the slot width is .804" provided by a reground 7/8 end mill which was the closest thing to a
20 mm in the shop. While I touched of the mill to go no deeper than .002" than the existing floor of the channel the
extra length broke through the "V" roof which was raised .025" and widened .250" to drop the saddle earlier to get
the cross slide dove rails level to bed. I could have gone back further as other have but I want to keep the tool inside
the rails. I will epoxy a disc in this location to prevent chips from entering the "V" channel as this become exposed
when the slide is extended with the bearing modifications.
Note the even oil film on the back way? Full and flat contact! Fully adjusted this saddle will slide right off the end of
the bed and go right back on without a hitch with even light drag all the way to the head stock area even though it's fit
to near zero gap.
measurement for shim thickness is above each but a bit of fine tuning on a sheet of 2000 grit with thinning in
increments of .0001" per fitting was required to get it right. Of course that means I had to make each one twice and
you over shoot the dimension one step the first pass. The different thickness is required as the saddle casting is
not parallel to the bed at their mounting face.
The six inch length was a vast improvement over the standard 3.9" stock and 4" X .352" thick version. The extra
length will preclude the use of the motor guard but I don't plan on the stock location for the motor anyway. The extra
width was pure laziness at first but now that it's there I see a saddle lock possibility and perhaps a thread mount for
a DTI or as it turned out a gib with four workable wear faces.
Note that the area for the cross slide screw is milled out in anticipation of a thrust bearing in the collar later. The
extra length is .250" and the slot width is .804" provided by a reground 7/8 end mill which was the closest thing to a
20 mm in the shop. While I touched of the mill to go no deeper than .002" than the existing floor of the channel the
extra length broke through the "V" roof which was raised .025" and widened .250" to drop the saddle earlier to get
the cross slide dove rails level to bed. I could have gone back further as other have but I want to keep the tool inside
the rails. I will epoxy a disc in this location to prevent chips from entering the "V" channel as this become exposed
when the slide is extended with the bearing modifications.
Note the even oil film on the back way? Full and flat contact! Fully adjusted this saddle will slide right off the end of
the bed and go right back on without a hitch with even light drag all the way to the head stock area even though it's fit
to near zero gap.
Several notes in this view. Note how close I got the gib width to the bed casting? It's all about stiffness and
surface area, especially on the rear gib. I will shorten the stand offs latter now that I have the shim pack sorted and
add some thin washers under the jamb nuts as I did the button heads.
The material is A-2 but I left it unhardened unlike the start up plan. I can go back and do that latter if it becomes
a problem but for now the thickness is preventing the bowing and twisting so apparent in the factory gib
arrangement. Sooner or latter they will get treated to at least a flash hard chroming for corrosion resistance. Both
the front and back gibs are ground both sides to .00005 flat and parallel at a 32 RMS finish then lapped lightly with
1500 grit to take the tops off the asperities to aid break in and provide oil retention. The other four surfaces are just
milled to .001 for square, parallel and perpendicular to the faces. This gib is wide enough to be reversed and
inverted providing four wear faces if needed and is between 20 and 25 HR-C hard. All edges are chamfered 1/32 x
45 degrees. It can also be reground several times so it's pretty much a life time gib.
This geometry provides a 150% increase in surface area over the absolute best possible fitments with the
factory gibs or even current tapered gibs and 24 X what was actually contacting when this started. The factory gib
can be bent with your hands. This one will take several hundred pounds force to deflect .001". The casting will
break first. ( Don't worry, there isn't enough power to break the bed)
I noted earlier that 260 brass shims would dent the factory sintered iron gibs. 260 brass is about Rockwell B
scale 77 hard or 140 Vickers. Annealed A-2 is about Rockwell C 20-25 or 239-266 Vickers, not quite twice as hard.
However for those riding the hard will wear the bed bus G-2 Class 40 gray iron is between 207 to 283 Vickers
before hardening and can be hardened to about Rockwell C 45 or Vickers 447. I'd say A-2 is a fair match for this
bed and the gib should wear first and much slower than the sintered iron factory pieces (or brass replacements),
which by default, are something less than the 140 Vickers reading of hobby brass that deforms them.
surface area, especially on the rear gib. I will shorten the stand offs latter now that I have the shim pack sorted and
add some thin washers under the jamb nuts as I did the button heads.
The material is A-2 but I left it unhardened unlike the start up plan. I can go back and do that latter if it becomes
a problem but for now the thickness is preventing the bowing and twisting so apparent in the factory gib
arrangement. Sooner or latter they will get treated to at least a flash hard chroming for corrosion resistance. Both
the front and back gibs are ground both sides to .00005 flat and parallel at a 32 RMS finish then lapped lightly with
1500 grit to take the tops off the asperities to aid break in and provide oil retention. The other four surfaces are just
milled to .001 for square, parallel and perpendicular to the faces. This gib is wide enough to be reversed and
inverted providing four wear faces if needed and is between 20 and 25 HR-C hard. All edges are chamfered 1/32 x
45 degrees. It can also be reground several times so it's pretty much a life time gib.
This geometry provides a 150% increase in surface area over the absolute best possible fitments with the
factory gibs or even current tapered gibs and 24 X what was actually contacting when this started. The factory gib
can be bent with your hands. This one will take several hundred pounds force to deflect .001". The casting will
break first. ( Don't worry, there isn't enough power to break the bed)
I noted earlier that 260 brass shims would dent the factory sintered iron gibs. 260 brass is about Rockwell B
scale 77 hard or 140 Vickers. Annealed A-2 is about Rockwell C 20-25 or 239-266 Vickers, not quite twice as hard.
However for those riding the hard will wear the bed bus G-2 Class 40 gray iron is between 207 to 283 Vickers
before hardening and can be hardened to about Rockwell C 45 or Vickers 447. I'd say A-2 is a fair match for this
bed and the gib should wear first and much slower than the sintered iron factory pieces (or brass replacements),
which by default, are something less than the 140 Vickers reading of hobby brass that deforms them.
Notes on the front gib include the decided lack of hardware, shorter dimension but it's just as wide as the
rear. It too shims on the outside edge only but full width and again unequal thickness. Without the full width you
get a different wear pattern every time you adjust them as they rock about the stand off screws. On the factory gib
these stand offs are not dimensionally anymore correct than anything else on the machine. They are not centered
nor in the direct bolt line plane. They have no choice but to rock about their axis. Not centered in either dimension
they induce bowing and twisting.
This gib is also A-2 and finished to the same standards for fit and finish as the rear. Unlike the rear it doesn't
have to deal with the same forces thus shorter and thinner at 4" in length and .352" in thickness. (I see I need
washers here as well). The extra width gib can't be done here (apron interference) and is unneeded as little wear
will occur at this location unless your working outside the bed way with some far over reaching tooling set up that
this width bed isn't designed for.
Before you ask, yes the apron will bolt up flush BUT the rack pinion will not rotate. Instead of creating a stress
riser by carving on it as the taper gib arrangement does, I will turn the excess width from the gear. It's about twice
as wide as the rack and the shaft proper does clear this fit buy .002"
Button heads again for a cleaner look and a bit more room. The stock caps are 6 mm X 12 mm at 1 mm pitch.
These front caps are 16 mm long and the rear ones 20 mm long with 1.5 mm thick washers. That may be useful,
right? The stand off on the rear gib are 1/4 X 1 at 20 TPI. Don't worry about the coarseness of the thread. You don't
use them in this set up for adjustment, only support and stiffness.
Now not everyone has access to a good surface grinder so for DIY types you can buy annealed A-2 bar stock
that is .750 wide and use the drawing above for bolt locations. The front one could be .250" thick and the rear .500"
thick. Just cut to length and drill away :)
The most important aspect of this project was get the underside flat and parallel to the top and use shims in
addition to fasteners to keep it parallel. If your really strapped for cash and must live the factory gibs at least
straighten them, lap them and use shims in addition to the cap screws and standoffs to keep bow, rock and twist
to a minimum.
When you get things this aligned and use a good lube it is just magic what happens. I wish there was some
way to show or demonstrate this on a web page.
rear. It too shims on the outside edge only but full width and again unequal thickness. Without the full width you
get a different wear pattern every time you adjust them as they rock about the stand off screws. On the factory gib
these stand offs are not dimensionally anymore correct than anything else on the machine. They are not centered
nor in the direct bolt line plane. They have no choice but to rock about their axis. Not centered in either dimension
they induce bowing and twisting.
This gib is also A-2 and finished to the same standards for fit and finish as the rear. Unlike the rear it doesn't
have to deal with the same forces thus shorter and thinner at 4" in length and .352" in thickness. (I see I need
washers here as well). The extra width gib can't be done here (apron interference) and is unneeded as little wear
will occur at this location unless your working outside the bed way with some far over reaching tooling set up that
this width bed isn't designed for.
Before you ask, yes the apron will bolt up flush BUT the rack pinion will not rotate. Instead of creating a stress
riser by carving on it as the taper gib arrangement does, I will turn the excess width from the gear. It's about twice
as wide as the rack and the shaft proper does clear this fit buy .002"
Button heads again for a cleaner look and a bit more room. The stock caps are 6 mm X 12 mm at 1 mm pitch.
These front caps are 16 mm long and the rear ones 20 mm long with 1.5 mm thick washers. That may be useful,
right? The stand off on the rear gib are 1/4 X 1 at 20 TPI. Don't worry about the coarseness of the thread. You don't
use them in this set up for adjustment, only support and stiffness.
Now not everyone has access to a good surface grinder so for DIY types you can buy annealed A-2 bar stock
that is .750 wide and use the drawing above for bolt locations. The front one could be .250" thick and the rear .500"
thick. Just cut to length and drill away :)
The most important aspect of this project was get the underside flat and parallel to the top and use shims in
addition to fasteners to keep it parallel. If your really strapped for cash and must live the factory gibs at least
straighten them, lap them and use shims in addition to the cap screws and standoffs to keep bow, rock and twist
to a minimum.
When you get things this aligned and use a good lube it is just magic what happens. I wish there was some
way to show or demonstrate this on a web page.