Westminster Style Folding Library Steps

Westminster Style Folding Library Steps
Matt Morse takes on a challenging project for a client’s library

Over the summer last year I was asked if I could make a set of steps for a client’s home library. Picturing some kind of step stool I said that: “Of course I could make some library steps; how hard could it be?”. Chris Schwarz remarked: “Careful, could be the Barnsley Steps!” While certainly not the Barnsley Steps (the unusual steps designed by James Ryan at the Edward Barnsley Workshop), this project was beyond a simple step stool and it included some compound angled through joints and mind-straining angles. Half an hour on Google and a field trip to see an example in an antique shop and I had an idea of what was required. The brief was: steps in the Westminster style, slightly stronger visually and literally, than the antique examples to allow access to floor-to-ceiling shelves up to 3m high. The Westminster style has two useful party tricks; first, there are handrails on both sides and second, they fold flat including the handrails. These tricks make them very useful in the modern setting as an elegant stepladder that can be folded away when not required.

Design
Breaking down the design is relatively simple. There is a ladder section with five treads, a back section to form the other main side to the A-frame, a small upright on each side and a handrail on each side. Complexity is introduced as the steps diminish in width towards the top and each tread is angled back at the same angle as the ladder. The angle of the sides to the floor was determined by drawing several variants to scale until it looked visually right. The angle of the base of the legs was finalised at 60° from horizontal. All the treads follow this angle so when the ladder is open the treads are parallel to the floor.  The tread heights were 250mm on centre making the whole ladder section 1500mm long. The material size for the ladder was 30mm thick by 125mm wide. Material for the rest of the structure was also 30mm thick to visually balance the design. Determining the length of each tread and the angle of the sides was problematic theoretically; at least for me! In the olden days I’d probably have drawn it full size on an 8×4 sheet of something and used it as a rod for the build. These days I tend to use SketchUp. The tape measure and protractor tools meant in very short time I had a cutting list and all the required angles. This method was more accurate and quicker than full-scale drawing and gives a 3D model to play with.

Once the angles were established the next step was to determine the correct pivot points to allow the steps to support weight and fold easily. Initially several scale models were made from scrap. Then a full-size mock-up in softwood was quickly cut out and screwed together. This proved the size and proportions were correct but more importantly allowed the mechanism to be tested at full scale. The steps could be scaled up or down but the main principle for the folding mechanism is to keep the pivot points equidistant in all the uprights and in all the diagonals.

Design details
The hardware used was stainless M10 carriage bolts with dome nuts and nylon bearing washers. The stainless hardware was selected to allow polishing to a mirror shine without concern about removing any coating or plating. The dome nuts were designed to prevent catching clothing or users on the steps and gave a good visual finish. The rim of the bolts was counterbored just enough to make the 90° section disappear and to give the illusion that the bolt heads blended into the surface. Between the moving parts, 2mm nylon bearing washers were used. These gave a tight clearance and allowed the bolts to be relatively tight but still allow easy movement.

Building the ladder section
In this project 90% of the joinery involved is in the ladder section. Using the mock-up components as templates it was relatively simple to mark out and prepare the final stock from 32mm rough sawn boards.
The first challenge was encountered here when one of the two quartersawn boards for the ladder sides turned out after planing to be full of shakes making them unusable. The spare contingency stock was not quartersawn resulting in one quartersawn side and one crown cut side.
The base needs to be cut at 30° across the board with a 3° compound angle. This gives good bearing surface in the finished steps for stability. The top end was simply rounded over to a pleasing radius on the bandsaw and the disc sander. I normally find something roundabout the right size and trace round it to give a line to cut to. The base had its points rounded over in a similar way this time by tracing round a 13mm socket from the socket set.

The centre lines of each tread were laid out in pencil and a jig was used as a fence to guide the Domino. In my experience the simpler the jig, the more likely it is to work; this was a stick with some pencil marks on it and another stick as a shim. Lining up the pencil marks with the layout put the Domino in the right place for the first run of mortises, using the shim offset the Domino to give the final width of the mortise. The two ends were mortised right through to a clamped-on backer board to prevent blowout on the face side. The centre web of each mortise was cut to 10mm depth only; enough to give full support to the tread but not detract from the strength in the side. All the through mortises were 25mm wide, 30mm long and separated by 35mm from the other mortise for that tread. There was a 5mm shoulder on all sides of the tread’s tenon.
The mortises were cut at 3º to accommodate the 3º angle of the sides of the steps. The Domino was tried on a test cut set to 3º but in the test cuts it didn’t achieve a 3º mortise owing to the fence on the machine protruding slightly. It was probably not designed to operate this way but from trial and error it was found that by setting to 4º, a 3º mortise was achieved. Layout lines were then transferred and the mortises were squared using a chisel and gently flared towards the outside for wedging.
From the SketchUp model and the full size mock-up the shoulder lines for the treads were marked out and a generous tenon added before cutting to length. The tenons needed to be long enough to go through the mortises and leave some length to allow a chamfer. This is lost in the final clean-up of the outside but is invaluable during dry fit and final assembly to prevent blowing out the face sides of the mortises. The cheeks of the tenons were cut on the router table staying well clear of the shoulder lines. The shoulders were cut a few millimetres from the line with a handsaw taking care to preserve the 3º angle. The final few shavings were pared back to the line using a large guide block cut to 3º to help cut the upper and lower shoulders accurately at 3º. The shoulders are one of the structural elements that prevent racking and needed to be cut accurately. 

With the angled shoulders cut, the remaining parts of the tenons were cut with a handsaw and chisels. The individual width of each through tenon was determined by showing the tenon to the mortise and taking the sizes directly from each mating piece. No measuring for increased accuracy! Each tread had to be marked from the outside of the side and placed upside down for marking out. When turned over and put on the inside the sizes corresponded.
The treads were then individually dry fitted after fine-tuning the fit in all dimensions with router plane and shoulder plane. After a full dry fit the parts were smooth planed to remove all the layout line and the decorative details were cut.

At this point the holes for the hardware and the counterbores were cut and the square recesses for the square shank of the bolts were cut with a chisel. Inking the shank of the bolt with a marker and giving it a tap in the hole marks the cut perfectly. I aligned the square sides on the diagonal with the grain to allow easy cross grain paring for all sides of the hole.
The inside front edges of the ladder sides in the original examples I have seen are beaded. This gives a nice smooth edge with a decorative detail but also gives a shadow line. This can be an important visual cue as to where the edge is so I included this in the sides and on the treads. I cut the bead using an old wooden side bead plane. One side was cooperative and the other was not so I had to plane against the grain. It’s best to minimise the tear-out in some way when going against the grain so I used two cutting gauges to define the sides of the quirk to prevent surface tearing. I then rounded the edge with a block plane with the grain to minimise tear-out on the rounded section. Care is needed here not to round over too much or the beading plane will struggle to engage on the edge – because it’s been taken off!
All other edges including the feet and the rounded tops of the ladder sides were then rounded over using a similar radius round-over bit in a hand-held electric router. The exposed edges of the treads were all beaded with a slightly smaller beading plane. The quirk from the beading planes can be a little sharp so I used the tiny Veritas shoulder plane to bevel this back on all the beads.
All the rounds and surfaces were then blended together using Abranet on a sanding block and the pieces were sanded to 320 grit by hand for pre-finishing. Sanding sealer followed by multiple coats of shellac applied by rubber was used to give a prefinished surface before assembly. 

Wedging the tenons for the treads
The tenons for the treads needed some mechanical lock for maximum strength in this part of the steps that would take the user’s weight and be subject to racking forces from side to side. Normally wedges go across the tenon and across the grain of the mortise component. In this case the two parts were at an angle and so diagonally across the tenon was close to straight across the grain of the side piece. This has the advantage of being able to flare the tenon in two directions making it tight with all walls of the mortise. It has the disadvantage that the wedges need to be shaped to fit the corners of the mortises and the wedges themselves are at a 3° taper angle! After an hour freehanding with a block plane and achieving only one successful wedge and several shortened fingertips, I thought some sort of fixture might be in order. Using the chopsaw I cut two blocks for the shooting board at compound angles – 3° in the width and 87° and 93° for the length. I used these like a mitre shooting board to accurately put the correct angles on one end before marking the length and putting the angles on the other. The fixtures needed to be recut with the 3° the other way for the other side of the ladder.

Glue-up
My usual first choice of glue for oak (Quercus robur) is Titebond III as the dried glue is a reasonable colour match. The open time is only about 15 minutes though so a bit short for sinking 20 wedges in the five treads! I opted instead for liquid hide glue, which has about double the open time. I warmed the glue in warm water for 10 minutes before use to make it flow and soak into the wood better. As these mortises are cut on the diagonal the majority of glue surface in the mortise is end grain. I applied a size coat for all the end grain surfaces before going back five minutes later with a second proper coat. The size coat fills the pores and allows the glue coat to take hold, rather than wick away down the fibres.
The tenons had all been fettled to give an easy sliding fit and the assembly came together reasonably well with only a few taps from the rubber mallet. Before assembly I’d made up some angled clamp heads and taped them to the clamps to allow clamping of the sides without bruising the timber. With the sides clamped tight and all the mortises completely closed the wedges were glued and hammered in to the same depth in each mortise.
After the glue had set the tenons were trimmed with a flushcut saw and the sides were planed smooth. Finishing was left until after the other moving sections had been test fitted.

Building the handrails, uprights and back section
The small upright sections and the diagonal handrails are simply rounded sections with holes and counterbores at the appropriate locations. The ends were rounded in a similar way to the top of the ladder section with a bandsaw and disc sander. A pleasing oval cross section was achieved using a thumbnail profile bit in the router table followed by blending of the surfaces with hand sanding.
The back support is where the remaining joinery is and first of all the pieces were roughly shaped leaving a little overlength at the base for fine-tuning. The top sections were taken to the oval profile as for the handrail sections above and the lower sections given the same roundover as the ladder section. The areas to be jointed were left square at this point. A test assembly of the whole piece determined the final length and angle of the back piece. By shimming the main ladder up by 6mm and levelling the whole structure it was possible to scribe the angle and length on the back pieces using a 6mm spacer.
A simple cross halving joint was cut on the two oak battens for the rear brace. This assembly was dry fitted and clamped to the ladder section to mark out the shoulders for the joints. The shape of the dovetail was marked out in pencil and cut with a tapered wedge on the bandsaw. The bandsaw was chosen as it’s difficult to start a cut with a handsaw on a corner. The pieces were cut back on the bandsaw to reduce the thickness of the dovetail to half the thickness of the stock. This was to maximise strength in the rear uprights. The dovetails were cleaned up with chisels and router plane ready for marking out the sockets. The acute angle on the inside of the tails was cleaned out using a combination of knife, saw and the raker tooth on the end of a Japanese pull saw. The cross brace was dry assembled again and laid onto the back section of the steps for marking the dovetail sockets. The components were prefinished as far as possible prior to final assembly to give an even finish in the tight angles of the brace section.

Final assembly
After final sanding and polishing all the components one dry fit was needed to determine the locations of the stops. These prevent the steps from closing up in use and the location is critical to allow it to open and close but stop at the required angle. These were located, cut and shaped and jointed with two 14mm Dominoes each. Clearance cutouts were needed in the back brace to give maximum bearing on the stops and allow clearance for folding. These were cut with a crosscut saw and chisels and refined with abrasives.

Conclusion
Making these steps pushed the boundaries a little for me with the angled mortises and diagonal wedging. The models and the prototyping made it a similar project to designing a chair – the first one takes all the time! Next time I would make sure I had quartersawn stock for the main components for a better visual match. I’d also take better care of the dovetails and sockets as the very delicate points on two of them became softened with handling. Overall a success and delivered on the day of the deadline in time for a surprise 70th birthday present!

0%

0%

Awesome

• User Ratings (2 Votes)
  6.8