Fort Vancouver bench

Patrick Neal forwarded some detailed pictures of the bench at Fort Vancouver, which was apparently constructed to Nicholson's original specifications.  Here they are:

Patrick says "[t]he bench is 32" high, the top is 30 x 104.5, the legs are true 4 x 4, the stretchers are also 4 x 4, and the skirt and top support boards are nearly a true 2 x 6.  He says the vise needs a guide, demonstrating the wisdom of what Mike Siemsen has done (see below).

Jamie Bacon's Bench

Here is a picture of Jamie Bacon's bench (used by permission).  Read about it on his blog:  Plane Shavings.  Notice that he put notches on the legs into the side board, as in the original engraving.   Jamie told me that he  really likes his front vise "but if I had it to do over, I’d have just made a portable one like the Moxon that everybody’s gone nuts about this past year."  I really like the clean look of not having a front stretcher or shelf, although Jamie has since added a shelf to his.

Southern yellow pine is a nice wood for this bench.  Wish we could get it here.

Fort Vancouver National Historic Site Carpenter's Shop

Patrick Neal comments below that he volunteers at the Fort Vancouver National Historic Site Carpenters' Shop where there are two Nicholson benches built from Nicholson's instructions.  I spent some time tonight searching for photos and found the following panorama, which is copyrighted so I'll just link to it here.  Carpenter's Shop Panorama  Read about the photographer here:  Don Bain.  The picture is quite unusual in that you can click on it with your mouse and look around the shop.

Notice that the benches only have the front sideboard, which is authentic.  However, notice the vise on the bench to the left against the wall.  Do you see what's on the guide?  Looks to me that it functions like a horizontal leg vise with holes and a peg in the guide.  Is it the functional equivalent of a twin screw vise?

I'm definitely going to go for a visit soon, take some detailed pictures to post here and try to learn more about their historical research.   Thank you Patrick.

Mass and rigidity

Nicholson benches are sometimes criticized as lacking mass and rigidity because the top is relatively thin.  One of Christopher Schwarz's rule is to always add mass.  My bench will contain roughly 111 board feet of lumber and will be 8 feet long by 2 feet wide.  Assuming a Roubo with the same surface area, a 5 inch thick top, the same size legs and 2 inch by 6 inch stretchers around the bottom, it would contain roughly 122 board feet of material.  To make things almost equal, assume I put a transverse bearer every foot across the Nicholson bench, bringing the total up to 117 board feet.

These benches have roughly the same mass and the same work area.  Which is more rigid?  Not being a structural engineer, I don't know the answer, though I want to explore the question at some point if I can.  Intuitively, there is actually more structure in the top of the Nicholson and the side boards take advantage of the properties of wood to resist stretching and compression along the grain.  Torsion boxes are known for their rigidity.  I can tell you for sure that a 2 inch thick fir board supported on 1 foot centers is pretty solid.  Of course, so is a 5 inch thick plank supported on 6 foot centers.  Any engineers care to comment?

Until proven otherwise, I think both of these benches have more than adequate rigidity.  That's probably why both traditions came about.  Different ways of skinning the cat.  My point here is that there is no reason that a Nicholson bench need have less mass or less rigidity than the alternatives.  That's a design choice.

Mike Siemsen's bench

Mike Siemsen has a woodworking school north of Minneapolis and a website.  Three years ago, he began posting about his construction of a Nicholson bench on his blog (Mike's blog).  This is how he characterizes his requirements:

I needed a bench that was portable and fit in my van to take to demonstrations. I wanted it to be an inexpensive example of a workbench for people getting into woodworking without a large outlay of cash. I wanted a bench that could be easily switched from right handed to left handed use. I wanted to build my own wooden vises. It needed to be stout.

He spends about $100 on his benches using construction materials and, as you can see, ends up with a very nice bench.  I learned a lot from reading his blog posts.  One thing of particular interest to me is the way he constructed his vise, which is quite similar to the one in the original engraving.  He made an assembly consisting of a tube for the guide to run in and also to hold the screw securely (picture here).  There has been criticism of the original vise and this seems like it would take care of it.  I emailed Mike about this and his opinion is that all vises rack and you need to use a spacer block.  Before deciding not to include a vise, I had been musing about modifying the original vise to operate like a leg vise installed horizontally, with holes and a pin in the guide.  I think Mike's solution is superior.

Here's a picture of the completed bench from this page of his blog.

Mike also shared the news that he is hoping to post an article around Christmas time about a "simplified" version of the bench.  It took him 9 hours and cost $110 to build, including a steel screw for the vise that he bought on sale at Lee Valley!  Even though I chose to spend more money so I wouldn't have to use construction lumber, I really admire his approach.  As Mike says, "These benches are very simple and we tend to over think things."

You can make a Nicholson bench as fancy and expensive as you want, but it is good to remember how simple and economical they can be.

Update:  Someone asked Mike about the traditional vise on his bench and he gave a thoughtful reply on his blog.  Read it here.   

Bob Rozaieski's bench

As many of you know, Bob has a popular website called  Logan Cabinet Shoppe .   A year and a half ago, Bob began a series of podcasts in which he constructs a Nicholson bench using handtools without using an existing bench.  This series is really interesting and a must view for anyone considering building one of these benches.  I just want to focus here on one episode that has mostly to do with design and use, Episode 23, Workbench Work Holding.

One of the most important attributes of a workbench designed for hand tools is workholding.  Others are mass and rigidity, which I'll discuss separately.  When you use stationary machines, you take the piece to the machine.  When you use hand tools you take the tool to the piece, usually on a bench.  It is crucial that the piece be held firmly and in a way that allows the operation to be performed conveniently and effectively.  Ideally, you'd position the piece on the bench just the way you want to work on it, snap your finger and it would be instantly locked into place.

A popular way is to do this with vises, which have attained limitless variety and an almost unimaginable degree of quality and sophistication.  For example, Benchcrafted sells a "Roubo Benchmaker's Package" consisting primarily of exquisite vise hardware for $750!  A more traditional alternative is Lake Erie Toolworks, which sells a fantastic premium wooden screw kit to make a twin screw vise for $365 or you can buy one screw for other alternatives like leg vises.  Lee Valley and Lie Nielsen both make hardware for twin screw vises.  These are amazing products.  Of course, there are more economical alternatives as well.

For Bob, simpler is better.  His bench has a basic twin wooden screw vise attached to the side board that is often not even on the bench.  That's it.  All other work holding is accomplished in other ways.  Episode 23 is a demonstration of those ways.  For me, it was an eye opener and fundamentally changed my way of thinking.  Watch it to see a minimalist, highly effective traditionalist alternative.  It comes down to having a system of stops that oppose the force you are applying.  What Nicholson calls a bench hook and we call a planing stop holds a piece from moving side to side (the x axis).  The split top accepts a stop that runs the length of the bench in the middle and holds a piece from moving front to back (the y axis).  Holdfasts hold the piece from moving up and down (the z axis).  It's that simple.

Similar thinking applies to pieces held vertically on the front side board.  The board is held up by dogs or holdfasts.  It is held to the sideboard with a "crochet,"  a piece of wood shaped like a finger holding the end of the board, and holdfasts.  That's it.

The vise still has a place, most notably to hold pieces vertically for dovetailing, sawing tenons, etc.

You either like this or you don't and there is a not a right answer.  I am as impressed as most by the benches with these beautiful vises.  But, this simple approach appeals to me for its historical authenticity, its elegance and its economy.  As best I could, I experimented with my existing bench and found I really liked it.  To me, this episode makes an overwhelming case for a Nicholson bench that employs traditional work holding methods.  I have decided to go one step further as an experiment.  My bench is not going to have a vise at all, though I did think through how I will add one later if I choose to.

Notice that I didn't say I wouldn't have a vise.  I'll post about the Moxon vise, which I will have, separately, but the point is it won't be fixed to the bench.  And it might use hardware from one of these companies.  I'm not as traditional, or as accomplished, as Bob.

Building begins

My lumber is S2S 1 15/16" thick so I made my legs 4x6s.  Since it's dead flat, this was super simple.

Next, I started working on the side boards.  Although it isn't in the Nicholson engraving, a number of modern builders have cut off the corners and I like the look.  I wanted to get a nice, clean finish cut, so I adopted the somewhat unorthodox approach of using my big Disston miter box saw.  It worked great:

I am trying to be very careful to make sure the top boards rest on a perfectly flat plane (among other things, to minimize the effort of flattening the top) so I took a lot of time matching up the side boards exactly:

Laying out the holes on the side boards

I have been asked about laying out the holes on the side boards.  Assuming that the legs on the bench in the engraving are 10 feet apart on center, each set of holes is 2 feet wide.  The vertical row of holes continues down the legs.  The actual width of a 2x12 is 11 1/4 inches.  If you make the vertical distance between the centers of the top and bottom rows of holes 7 inches, then the center of the holes on the diagonal will be 5 inches apart.

As long as the center of your legs is an even multiple of 2 feet apart, this will work.  For an 8 foot bench with legs 6 feet on center there will be 3 sets of holes, for a 6 foot bench with legs 4 feet on center, there will be 2 sets of holes, etc.

7 inches makes the math works out evenly but, if you want other dimensions, just remember the Pythagorean Theorem.  Add the square of the horizontal distance and the square of the vertical distance of each set of holes.  Then take the square root of that number.  That's the length of the diagonal.  Divide by 5 to get the distance between holes.  In my example, 7 squared plus 24 squared equals 625.  The square root of 625 is 25.  25 divided by 5 is 5 inches between the centers of the holes measured along the diagonal.  Of course, you can alter the number of holes as well.

If you'd like the vertical distance between the top and bottom row of holes to be 9", then the holes will be 5.126" apart along the diagonal.  This is so close to 5 1/8" that I think we can call it good.  Just measure from the top or bottom consistently.

Similarly, if you want the vertical distance to be 8", then holes placed at 5 1/16" intervals along the diagonal will be almost even.  Just start from the same edge every time.  In the end, 8" looked about right to me, so that's what I did.

As for the vertical holes, since there are three of them, the easiest thing to do is use dividers.  That's probably how Nicholson would have laid out the holes on the diagonal too.  Really no need to use a ruler at all if you want to skip the math.

Wood Movement

One of the things I have wondered about is that several respected woodworkers have glued the side boards to the legs, a cross grain joint.  Because the side boards are so wide, shrinkage is a major issue, particularly here in Oregon where we have hot dry summers and cool wet winters.  According to the Woodweb calculator, and assuming the moisture content ranges from 7 to 14%, this could be as much as 5/32".  Perhaps extreme, but it convinces me to take shrinkage into account.

I have decided to deal with shrinkage by floating the torsion box on the legs.  I will glue and peg the bottom of the side boards to the legs and extend them 1/4" above the top of the legs.  The 6" transverse bearers will rest on a 6" ledger glued to the side boards, which will allow the sideboards to expand and contract freely without affecting the top.  This ledger also serves as a long stretcher and will keep the bench from racking.  As a result, my bench will be like the Nicholson plate in not having a long stretcher across the bottom.

Why I'm not using construction lumber for my bench

As Nicholson observed, the thicker the better.  A criticism sometimes made of the Nicholson bench is that it is springy and light.  I think this can be addressed by using full 2" pieces and by stiffening the torsion box.  After using his Nicholson bench for a year, Bob Rozaieski said that, if he had it to do over again, he would make the top a full 2" so holdfasts would work without gluing blocks to the underside of the top.  A good compromise would be to make the base out of construction lumber and just make the top from full 2" higher quality lumber.  But, I decided that I would either use reclaimed lumber, which is often available in my area, or go to a custom sawmill for all the material.  As it happened, I stumbled onto what I regard as a great option.  A custom sawmill in my area had some kiln-dried almost clear vertical grain Douglas Fir that had been rejected for millwork because of stains.  I got it for less than $3 per board foot, much more than construction grade lumber but worth it in my opinion.  The pieces are dead flat, sanded and a full 2 inches thick.  Because they are vertical grain, I hope that the bench top will be very stable.  And. ...I like the way it looks. This lumber says Oregon.  It's acclimating in my shop now.  Time to settle on a design.  

What Nicholson had to say about the Bench

Only two pages (87-89) are devoted to the bench in Nicholson's book.  There are some interesting tidbits beyond what is apparent from the engraving.  These caught my attention.

Nicholson describes a "upright rectangular prismatic pin," visible in the engraving on the left side of the top.  Used as a planing stop, he calls it a bench hook.  It is essentially a square block held in place by friction that is moved up and down with a mallet to serve as a planing stop.  This is for a right-handed woodworker.  The bench hook and the screw would be on the right side of the bench for a left-handed woodworker like me.

The bench screw, or vise, has a "guide" on the side opposite the screw itself.  It is mortised through the side board and serves to align the top of the check with the top of the bench.  This vise is subject to racking unless a  spacer is placed between the check and sideboard on the opposite side of the screw from the piece being held.   Some modern variations of the bench use a twin screw vise instead of the guide.

The recommended dimensions of bench components fits reasonably well with standard dimensions of modern construction lumber.  The legs are to be at least three and a half inches square and the top is to be one and a half to two inches thick, "the thicker, the better for the work."   The latter is the major issue that arises in using construction lumber for the bench.

The pairs of legs are "generally coupled  together by two rails dovetailed into the legs."  The engraving looks like a mortise.  There are three or four transverse bearers between the legs.

" For the convenience of putting things out of the way, the rails at the ends are covered with boards."  He goes on to describe a locker that is sometimes created for tool storage, accessible by sliding a portion of the top open.  I don't see this in the engraving.

The Nicholson Bench

I decided to create this blog to document my construction of a Nicholson woodworking bench from the English woodworking tradition.  It appears in an early 19th century book called The Mechanic's Companion by Peter Nicholson.  (available online at Google Books).  The bench is shown in an engraving (Plate XII):

This workbench definitely doesn't look like what we expect.  Its distinguishing feature is the wide (approximately 12") "side board."  It is used to hold boards horizontally for operations such as edge planing.  One end of the board is held in the vise and it is held up by dogs or holdfasts placed in the holes at an appropriate height.  The side board also functions as a large rear "check" for the "bench screw."

The side board serves a structural purpose as well.  The top of the bench is relatively thin, about 2".  The side board makes up for this, transferring the force of planing, chopping etc. anywhere on the top to the legs and floor.  It also serves to anchor what Nicholson calls "transverse bearers" that support the bench across its width like floor joists.  It's  not apparent from the engraving, but from the text there is apparently a narrower side board along the back of the bench to hold the other end of the transverse bearers.  The result is what we call a torsion box.  Clever design substitutes for mass in this bench.  Perhaps a reason is that it is a joiners' bench intended to be transported to worksites rather than remaining in a workshop.

The dimensions of the bench are given as 10 to 12 feet long, about 2 feet 8 inches tall and about 2 feet 6 inches wide.