Flattening a bench top
Paul Pedersen
The following describes how I went about flattening the top
of my workbench. The top is 42" deep x 66" wide x 1-1/2" thick birch.
It had developed about a 1/8-1/4" dip in the center towards the front of
the bench.
My method involves the use of a router-carrying beam riding on rails
attached to each end of the bench. This setup is used only to define
the final plane of the top, not to remove all the wood.
I first made sure that all the lag bolts that attached the top to the
base were snug, and that the table was roughly level.
I then made a U-shaped beam out of 3/4" construction cedar (what I had
lying around). This beam must be at least the length of a diagonal of
the top. The inside width of the beam was such that my router could fit
(narrow dimension). The height of the sides of the beam are arbitrary,
mine were about 2-1/2", the result of ripping a 1x6 board in two. I
squared up all the pieces and glued them up. I then jointed the bottom
of the beam and sent it through the planer so the sides would be
parallel (height-wise). An important thing to note here is that the
inside face of the bottom is no longer parallel to anything.
Next, I made a router carriage out of 3/4" plywood consisting of two
T-shaped pieces screwed to a bottom that would fit inside the beam and
ride on the top edge of the sides :
||||||||||||||||||||| ||| |||
||||||||||||||||||||| ||| |||
--- ||||||||||| --- -- ||| ||| --
--- ||||||||||| --- ||| (router) |||
--- ||||||||||| --- ||| |||
--- ||||||||||| --- ||| |||
--- XXXXXXXXXXX --- XXXXXXXXXXXXXXXXXXXXXXXXX
--- XXXXXXXXXXX --- XXXXXXXXXXXXXXXXXXXXXXXXX
------------------- --------------------------------
------------------- --------------------------------
(end view) (side view)
I then (using the router carriage) routed 1/2" wide slots in the bottom
of the beam 12" long with 3" between slots (so the beam wouldn't fall
apart). Since the carriage has four contact points with the beam they
must be adjusted carefully so that the carriage doesn't rock. [Waxing
the top edge of the sides of the beam let the carriage slide easily].
I tacked one (straight, 3/4 x 3) rail to each end of the bench top,
leaving them about an 1/8" proud of the surface. Since these rails
define the final plane of the top, it is important that they be set as
coplanar as possible. I used a 48" level to roughly align them.
Using a 1/8" straight bit in the router, two diagonals are cut in the
top of the bench. The depth should be a bit more than the maximum dip
of the top. If the rails are adjusted properly, the depths of the
diagonals where they intersect will be equal. If not, the rails should
be adjusted.
I lightly clamped the beam at each end to the bench so it wouldn't move.
Since the beam is quite flexible it must be shimmed at 3 or 4 places
along its length using pairs of cedar shingles. I suggest the shingles
be taped with masking tape to ensure they don't move while routing. I
had them move on me during one router pass, the result being a cut of
about 1/8" deeper than the rest of the cuts.
The rails should be firmly attached once they are adjusted. It's then a
matter of routing a bunch of (1/8") lines parallel to the long dimension
of the bench every 3" or so from front to back. I offset the beam for
every second line so that the gaps which aren't routed were staggered.
OK, now the bench has a bunch of lines the depths of which are all
co-planar. My original intention was to hand plane down to the bottom
of each line (I feel the Neanderthal waking in me ;-) Well, after
planing for about an hour with a freshly tuned #5 Record jack plane I
came to the conclusion that cave-men must have been built like
line-backers. I just don't have the muscles or the endurance to plane a
large surface of birch. Aching all over, I searched for a better way...
I ended up using a dovetail bit (the only one I had that was
flat-bottomed) in my router (1hp Elu MOF96). I divided the top into 3"
x 3" squares and routed out to the depth of the least deep line
bordering each square (a fine adjustment knob is invaluable here). It
is very important to leave small bands of the original surface for the
router to ride on. These bands can be 1/8" or thinner once you get the
hang of it.
The result was a surface roughly planar made up of hundreds of 'tiles'
bordered by thin strips of original bench top. These were then planed
down by hand (easy) using the #5 followed by a #4. This part did
involve removing some minor humps, checking continuously with my
reference rule.
I decided to take the whole surface down below the depth of the line cut
when the shims moved. I considered just filling it (after all, it's
just a workbench) but then I'd have to look at it all the time and
besides, the bench is one of the few things I have at home that I can
show off to potential customers. This time went a lot faster, instead
of tiles I routed continuous 3" wide bands length-wise leaving about
1/8" between bands, and all at the same depth setting. Had to do the
planing all over again too, though.
Finally, I sanded the whole thing using my PC505 and oiled the top with
tung oil. As I mentioned in a previous post, the top is now within
0.01" of flat, with respect to a 48" reference rule.
A final note : my router now makes a pretty strange noise (like dragging
a shovel across a concrete floor :-( which I haven't gotten around to
checking, or fixing. All in all, I'd say that it ran continuously for
two periods of 3-5 hours each. I don't know what 'production' use means
with respect to hours of continuous operation, but I'm not impressed
with this latest performance of my Elu. By the way, I had to sharpen my
carbide dovetail bit maybe three times using a 400 grit Grit-Bit (sort
of a toothbrush with a silicon-boron stone instead of the brush).
Paul
.
--
|GatorNet: Paul Pedersen 128:200/0
|Internet: Paul.P...@deltacom.cam.org
|
Walter Barry
>
>The following describes how I went about flattening the top
>of my workbench. The top is 42" deep x 66" wide x 1-1/2" thick birch.
>It had developed about a 1/8-1/4" dip in the center towards the front of
>the bench.
>My method involves the use of a router-carrying beam riding on rails
>attached to each end of the bench. This setup is used only to define
>the final plane of the top, not to remove all the wood.
I've seen this method before in FWW and a guy across the street from me
used it to do a VERY large table for a customer.
>...bordered by thin strips of original bench top. These were then planed
>down by hand (easy) using the #5 followed by a #4.
Actually, I would have used a #7 or (better yet) #8 to do the whole
job on the bench. If you really had some area which required a lot of
removal, mark them and use a scrub plane or #5 w/its blade rounded to
take deep 1/16" to 3/32" cuts. My bench is 3 1/2" thick Doug.Fir.
>A final note : my router now makes a pretty strange noise (like dragging
>a shovel across a concrete floor :-( which I haven't gotten around to
>checking, or fixing. All in all, I'd say that it ran continuously for
>two periods of 3-5 hours each.
Nothing against power tools but thats around 10hrs with setting up the
rails/bed right? Plus you mentioned yusing a plane anyway, and a
power sander. I used a PC352 belt sander on a bench once (outdoors)
along time ago... to long & to much mess. I own one now and wouldn't
use it for that again. This is one area where a hand jointer really
excels, and would do the job in less time, and although it would be
a really good workout, its less painful than listening to a router
for 10hrs.
Walter
Jon R. Bryan
: Nothing against power tools but thats around 10hrs with setting up the
: rails/bed right? Plus you mentioned yusing a plane anyway, and a
: power sander. I used a PC352 belt sander on a bench once (outdoors)
: along time ago... to long & to much mess. I own one now and wouldn't
: use it for that again. This is one area where a hand jointer really
: excels, and would do the job in less time, and although it would be
: a really good workout, its less painful than listening to a router
: for 10hrs.
I flattened the maple top of my bench with a #5 plane, a good straight
edge, a couple of winding sticks and a hand scraper. Took about two
hours, and the top had a fairly pronounced crown of about 1/16" in the
center. I am not a "Neanderthal," but hand planes sure sound neat when
they're sharp and cutting well :).
I had to experiment on the highest spots to find the best planing
technique for roughing down the surface. What worked best for me was
cutting across the grain at about a 45-degree angle, sometimes with the
plane parallel to the grain. In a couple of places I had to turn
the plane around and cut the other way to minimize tear-out. I then
took a much finer shaving with the grain for final smoothing and
finished off with the scraper.
The experience hasn't inspired me to get rid of any of my power tools,
but I do like Hock plane blades :).
--
Jon R. Bryan 505-844-2015 <=> br...@isrc.sandia.gov
Intelligent System Sensors and Controls Dept. 2111
Sandia National Laboratories
P.O. Box 5800, Albuquerque, New Mexico 87185-0949