Making Bolt Hole Chamfers with an End Mill on your CNC.

I spent the last 10 years using CNCs without a tool changer. Whenever the job required a bit change normally I thought about how I might be able to do the job using just one bit and some creative use of the CNC software I was using. This is one example.

Projects that needed perimeter or internal chamfers were done most efficiently with a bit change to a V bit.  I have a project that is put together with flat head bolts. The need to chamfer the bolt holes could have been done using a drill press after the CNC was done. Figuring out how to chamfer a hole with the same end mill I used for all the other cuts turned a 2 process job into one CNC  job with no bit changes. 

Vectric.com's Aspire and VCarve programs have a fluting toolpath.  Fluting takes a vector and ramps the bit down or down then up with the bit being centered on the vector. A circular array of 9 (or more) short vectors centered over the bolt hole, using a bit no larger than the diameter of the hole, can effectively chamfer the perimeter of the hole from the bolt head diameter down to the hole edge.  Using a small bit (1/8" or 1/16") may require a greater number of vectors to result in a relatively smooth chamfered edge. The chamfers in this image were made with a 1/4" diameter end mill. 

2 x 2 Hole Array

The slope of the chamfer and the diameter can be easily changed.  Aspire/VCarve also have a circular array layout tool that will use one vector that spans the hole and make any number you specify of rotated copies around the same center. Using an odd number of vectors results in a circle of lines with alternating start points.  Alternating the start points reduces the travel time needed move to each successive vector.  

Vectors Used

The center of the bit used is what follows the line. The depth the bit will slope down is set when creating the fluting toolpath.  In the image above the shaded area is a shadow of the area the 1/4" straight bit travelled over.  The short vectors used are shown in red. The center circle was the hole for the 1/4" bolt shaft. Depth of the fluting chamfer cuts was 1/8".

I made the vectors for one bolt hole, then simply copied them to all the other bolt locations. 

You can find downloadable  .CRV3d and .CRV file of the sample above on Vectric's user forum: Vectric.com

Questions or comments are encouraged!

4D

 

A Domino Slot Template.

Using Festool's Domino floating tenons is usually an expensive upgrade for hobby woodworkers.  The machine to cut the slots can cost over $1000, and the tenons themselves are more costly than traditional dowels made for joining wood parts. 

There are other less expensive sources for the floating tenons.  Attempts have been made to build simpler machines that duplicate the slot cutting action of the Domino machine.

I design and build furniture, and have found that floating tenons come in handy for connecting parts.  I can't justify the cost of a Domino machine, but I do have a small CNC and with it I can make the slots for ready made tenons on most furniture parts.  For joining long boards together I've designed a simple template that can be used with a handheld plunge router using a template bushing. 

A Domino Slot Template

This template could be made using a router table.  I drew up the template for cutting out with my small CNC.  The slots in the sides are for screws that would hold wood guide strips.  The guide strips can be adjusted to center or offset the slot that will be cut by the template. The template can be clamped to the edge of your board to stay in place when being used.  

A 5/8" diameter bushing in a plunge router base used with a 3/16"d spiral upcut router bit is all that is needed to make perfect sized slots for commercial Domino brand tenons or alternatives.  Some plunge bases for routers come with a dust collector shroud to suck away the chips as the slot is being cut. 

Plunge Router with Template Bushing

  This template is for 6mm x 20mm tenons. 40mm length. 

Slot has been Cut

You could make templates for each available size of tenon. All that needs changing is the size of the center hole. With floating tenons available in large quantity packs each size template would get plenty of use. The guide strips could be removed and used on whatever template plate is needed. A stack of different template plates would cost far less than the cheapest Domino machine you can find out there. 

A Good Fit

Once the template has been set up for the material, making several slots along the edge is relatively easy.  It just requires undoing the clamps, sliding the template to new position, then clamping it in place. Modifying the template for thicker or thinner material takes no longer to do than adjusting the fence of a Domino machine for the same job.

It is also easy to change the slot hole so you can use 5mm or 6mm bits to make the tenon slots.

For metric bits

If you use Vectric.com's VCarve or Aspire a files for these templates can be found on their forum HERE. 

Comments and Questions are encouraged!

4D

     

Tapering French (Sliding) Dovetail Joints for Ease of Assembly

Tapered French Dovetail 

French dovetail joints are a good way to connect critical cabinet parts.  A straight dovetail slot is easy to cut with a straight edge, router and a dovetail bit.  The mating male end of the joint takes a little more care to size right.  I've used a router table with the same dovetail bit and a tall fence to trim off the sides of the mating part end until the dovetail end fits snuggly into the slot. 

On wide boards a snug French dovetail joint can be difficult to slide in place. Friction builds up the farther into the slot you go. Getting the joint closed occasionally requires the pull of a bar clamp or tap of a wood mallet. The tendency of wood panels to bow or flex a little can add to the difficulty.  

Tapering the routed slot is fairly easy.  Make a straight slot against a straight edge first. Then move the leading end of the straight edge over 1/16" or so. Make note of how far you moved it.  A penny (1 cent) measures almost 1/16" thick (.06" rather than .0625"). I usually moved the straight edge end 1/16". 

To taper the male dovetail end a shim as thick as the amount the straight edge was moved is needed.  I tape a penny (tape + penny=1/16") or laminate sample chip to one edge of the mating piece just above the height of the dovetail bit. As the board slides past the router bit that shim holds one end away from the bit 1/16" to effectively taper that cut. The other side is cut parallel to the face. Care needs to be taken to match the side of the tapered slot to the tapered side of the joining piece. Cut on the wrong side the joint will still slide together, but be slightly crooked when in place.

CNC cut Dovetailed Side

CNC cut Tapered Slot

Using the CNC to cut the tapered slot is easy. The CNC cut tapered slot can be centered so there is no chance of ending up with a crooked panel. A vector that starts at the beginning of the slot, runs to the end of the slot, then returns to the beginning 1/16" away from the start point is all that is needed.  A profile toolpath centered on the line with cutting depth set for one pass for the dovetail bit will cut the tapered dovetail slot. 

Vector for Tapered Dovetail Slot (blue)

In the drawing above the green circles represent the diameter of the dovetail bit at beginning, middle, and end of the vector. The red lines show the width of the intersecting board. Blue is the vector needed and it is extended outside of the board by half the diameter of the bit.  This lets the bit drop down to cutting depth before it enters the edge of the board. To ease the stress on the dovetail bit I usually run a 1/4" down-cut spiral end mill on the same vector to clear out the slot before the dovetail bit runs to undercut it.

Vector for Male Dovetail End

Cutting the mating end requires being able to clamp the mating board vertically and level under the CNC spindle.   Making the toolpath to move the dovetail bit around the mating end requires some careful drafting.  You need to know the side angle, diameters of the dovetail bit, and depth of the dovetail cut.  Offsetting the male toolpath vector from the vector used to make the tapered slot will ensure a matching taper.  The amount you offset that vector is the necessary detail to come up with.  See the diagram above. One half of the bit tip diameter plus 1/2 the bit diameter at the cutting depth is how far you should offset the vector. I use a section view of the dovetail bit to find the cut depth diameter of the bit.  

Almost together. Still loose.

Alternately you could simplify the CNC toolpaths to duplicate the result of doing it without a CNC.  This is easiest if the dovetail isn't blind, and runs completely though the board. Two vector lines, with one slightly angled, to run the dovetail bit down on-the-line. To make the mating dovetail you'll need one pass down one side of the board, and an angled pass down the other side of he board. Placing the vectors for these requires knowing the bit tip diameter and the diameter at the cutting depth on the bit.  

With all CNC cut joinery the precision fit can be hard to nail down.  Using the same vectors and precise offset leaves no room for error or glue. I prefer to do the male side of the joint first, then the female slot.  I leave the slot side on the CNC to check the fit before unclamping it.  If the joint doesn't close completely you can reset your X axis (or Y axis depending on how you clamped up the board)  by a few thousandths and run the dovetail bit toolpath again.  I find .003" is a good amount of offset between sides of the joint for a good snug fit with room for glue. With sliding French dovetails it should take just a tap or two of a wood or plastic mallet to close the joint completely.  I tweaked my sample until my joint took two taps to close and will take the same two taps to free it up. 

Snug and Tight when completely together.

Comments and questions I encourage and welcome. 

4D

Tapered Halving Joint

Halving Joint

Half lap (halving) joints in standing wood parts can be made with many tools in a woodworking shop.  They are an obvious joint to use where flat pieces intersect and need to seem to pass through each other.

The disadvantage of the joint is when you want it snug you have to take care when cutting each side.  Yet when it is a very good fit the simple act of sliding the joint together can scrap off finish or mar the surface. 

I came up with a CNC cut version that is easy to slide together yet is very snug when assembled.  Both sides of the joint can be cut from the top side of the boards laying flat on the bed of the CNC.  I've used this joint several times between Baltic Birch plywood parts, and between two different thicknesses of hard maple for a student's project. 

Render from Aspire Software

One side of the slots cut has a tapered edge to slide against the tapered face of the opposing half.  Both sides are cut the same way. Care must be taken to match the slope of the face to the tapered edge of the slot.

Each Side Identical

When the thickness of the material is carefully measured, and the CNC toolpaths are done right this joint slides easily down until about 1/16" from closed.  A tap with a wood mallet drives it the last 1/16" for a joint so tight it'll take the mallet to open it up again. 

Almost Together

The ramped side helps the plywood stay perfectly aligned as it slides together.

A CNC file made for Vectric's Aspire or VCarve Pro/Desktop can be found HERE.  The slope made on the face of the boards was done with the fluting toolpath and a few parallel vectors.  I used a 3/16"d downcut spiral router bit for all the cuts.  If done with material thicker than 3/4" a  larger diameter bit may be needed.

Questions and Comments are welcomed and encouraged. 

4D

Rocking Chairs: Fundamental Design Considerations

When I taught furniture design to college students I was often asked how rocking chairs work, and how to determine the rocker design and radius. Over the years my students designed and built several successful rocking chairs.

Here are the basics:

The center of gravity (of the occupant + the chair) will be directly over where the rocker arc touches the floor when at rest. The distance from the floor to this center of gravity has to be shorter than the radius of the rocker arc. The closer these two values are together the farther the rocker will roll. The farther it rocks the longer the rocking period will be. Ideally you want to use a rocker radius that creates a rocking period equal to the occupant's at-rest breathing rate. You don't want the seat plane of the chair rocking forward past horizontal. 

Let M = distance from the floor straight up to the center of (the occupant's) mass above the floor plane (in inches).  On a more upright rocking chair design, the M may be 23" to 25". R = Radius (in inches) of the rocker arc.  39" is a good R value to start with.

Rocking 5 degrees forward and back

When M is less than R, or your center of mass is lower than the center of the rocker arc, the chair will return to center when rocked forward or back because the C.O.G. has been lifted up and gravity will pull it back down. In the diagram above the C.O.G lifts .0575" with 5 degrees of rocking forward or backwards.  

The center position is when M is directly below the center point of the rocker arc. From a side view, M is normally located near the occupant's belly button when he/she is in a sitting position. For women it may be slightly forward of the belly button. A value for M can be found by positioning the user above the floor in a comfortable at-rest sitting position. Measure from their belly button to the floor to find M.

Centered Position

R > M, and a good dimension for R is 39" or so. Increasing R (flattening the arc) reduces the period of a rocking cycle. Reducing R increases the period of a rocking cycle. By changing R slightly the rocking period can be "tuned" to match the breathing rate of the occupant. The simple act of breathing will shift the occupant's mass, causing the chair to rock back with each inhale and forward on exhale.

Having your feet flat on the floor will impede rocking motion. Ideally the occupant's feet should cycle from touching toes to touching down the heel as the chair rocks. This suggests the height of the front edge of the seat should be 16" to 18" or so. Less if the intended occupant is of small stature.  

This all works because shifting weight moves the occupant forward or back from being centered. Since the center of gravity is straight down from the center of the rocker arc, moving forward or back lifts the occupant uphill. The incline gets steeper the farther one rolls.  Shift your position on the rocker and you will "roll" to another centered position. Relax back to your rest position and the rocker slows to a stop.

If your rocking chair tried to throw you out when you rocked forward then the center of the rocker arc is too far backwards. If your rocking chair feels like you'll fall over backwards when you rock back then the center of the rocker arc is too far forward. I've seen and sat in student rocking chair designs that suffered from one or the other mistakes. 

Comments and questions are encouraged and welcome!

4D

Iterative Progression

In furniture or product design often what you hope will be a good design ends up with obvious room for improvement. This is why initial builds are considered to be prototypes. It takes seeing and testing the first prototype to realize where flaws exist or where there is room for improvement in aesthetics or strength or performance or functionality or simplicity of build.

Shown here is a sequence of Balans style chairs I designed and made. Inspired initially by the original rocking Balans chair my pursuit was to find a design that was simple to build, adjustable, and stable.  In my PhDesk article photos you can see most of an earlier 3 caster perch version done as a class project by my students. The design was static with no adjustability or flexibility. Link:  PhDesk Article

Imbuia Wood Collapsible 

My Imbuia and leather prototype above improves on that earlier student design with knee pads that could rotate to meet your shins at a whatever was the most comfortable angle.   Initially the frame post beneath the seat was intended to be moved to different positions along the lower rails. This prototype revealed that changing the angle of the seat would also tilt the caster stems off vertical and reduce the ease of rolling the chair around. The frame could collapse by lifting the center post off the pin it rests on. Collapsed it would  fit in a smaller box for shipping or storage.

3 wheels Adjustable Height

Highest Perch Position

The 3 wheeled version above could be adjusted in height/angle.  This design isolates the caster base from the adjustability of the seat and knee rest. An aluminum push button  releases the aluminum post when pushed in and locks the post position when released. Knee pads pivot to meet shins at the most comfortable angle.

4 Wheels Adjustable

In use.

While there is an economic benefit to using 3 casters rather that four, a 3 point footprint comes with a flaw discovered in use. They could tip and roll out from under the occupant when leaning to the back right or left.  This four wheeled version eliminated the tipping flaw of all the 3 wheeled versions.  This version stretched the frame back so the back caster beam was behind foot clearance. It had the same push button height adjustment and pivoting knee rest as the 3 wheeled version above.  

The sharp bend in the center frame of the 3 and 4 caster versions above required making them from 80 very thin veneer layers of wood. This later version used far fewer and thicker wood layers by changing the center frame to a smooth arc from under the seat down to the rear caster beam. Below a few photos of the arced frame. It has a seat that can be slid forward or back and locked in position with a cam lever.  The arc made room under the frame for occupant heels to meet or cross. This final version was gifted to the International Woodworking Fair management office in Fall 1988. 

Raised position

Lower position

The version above was the simplest build, the safest to sit on, and had adjustable height, seat position adjustment, and pivoting knee rests.  It was a design that only came about after making and using the previous designs. They were all built in a university fab lab and benefited from being tried out by several students and other professors. Feedback gained from each version led to advancements in later versions.  This sequence shows iterative progression in action.

An even later iteration I designed is my rocking Balans.  You can read about it HERE.   

Comments and questions are encouraged!

4D

Ze Chair. A Chair Design that Failed.

I designed and built Ze Chair in the Spring semester of 1980 while in college. It won first place in the quick assembly category of the 1980 IWF Design Emphasis Student Furniture Design Competition. With padded seat and back it was very comfortable to sit in. 

Ze Chair

"Ze Chair" was an obvious name for the zig zagging frame of this chair design. "Gravizy" was my second choice. It is gravity that holds the design in its Z shape. The design bolts together and can be taken apart for compact storage, packaging, and shipping. It can also be unfolded, although when stretched out flat the length is impractical to hang or store. 

The back of Z pivots to meet the back of the occupant at whatever angle is most comfortable. This meant that most who sat in it, no matter their posture, found it comfortable. 

I made two prototypes of this design.  The first had thinner parts and was made from locust.  The wood split at the bolted half lap joints when first put under load. This should have been a clue. I (temporally) solved that problem by increasing the part size and using red oak instead. My second prototype survived a competition (where it won 1st place) and 2 years of use before failing the same way the first prototype had.  It eventually started to split at the half lap corners. 

I also won an award for the poster design I designed for Ze Chair. It was a first place award from the National Association of Furniture Manufacturers, 1980 local student furniture design competition / poster division. Sadly I no longer have this poster or any record of it.   

I often revisit this design with the hope of finding a way to reproduce it without the potential flaws. I know now far more about joinery, the forces involved, and production methods for making the parts. The chair doesn't need to unfold, and that it did made it a challenge to pick up. That it can be quickly put together and taken apart is a nice feature though which I'd like to keep.

I've thought about trying this design one more time.  One idea is to trim away 3mm from each side of the half-lap corner joints and glue in a piece of 3mm Baltic birch plywood.  The plywood shouldn't split. Glued to the oak it should help keep the oak from splitting.  I can even pocket out a recess for the plywood so it doesn't show when the joint is together. First task would be to find/buy material to make the parts.  

If this was an initial design of a student of mine I would encourage them to keep iterating, perhaps toward some triangulation in the design that would remove splitting stress on the folding corners.   The bottom piece running from front to back could be eliminated and a rear leg could run up and connect to the angling front leg, seat rail and the arm rest for example. That would triangulate the structure and lock all the pieces together.

Questions and comments are encouraged!

4D

A Folding Desk

This is my PhDesk. I designed and built it in 1981. This original has a 12mm thick Baltic Birch plywood top, trimmed by 2" wide solid maple sides. The bottom panel is 1/4" thick hardboard. The front and back trim is rounded over to compliment the maple base details. It has a center pencil drawer that can be locked shut when the desk is folded up to carry.  The asymmetric X base folds up easily. The desk is 4" thick when folded flat.

PhDesk with Side Drawer

4" Thick when Folded

It was designed and built when I was a college student and moved several times. I wanted it to fold up so it would be easy to carry and would take up minimum space in the back of my small car. It won 1st place in the case goods category of the 1982 IWF Design Emphasis student furniture design competition. 

Several copies of this design were built for family and friends. My Furniture Design professor liked it enough to make his own copy of it while I was in the workshop making one.  Most have survived including my original prototype.  It was picked up by a furniture manufacturer who thought they could sell it to a line of kitchen accessory stores. They showed me a factory sample made from oak, but never succeeded in selling any.

Oak sample

My original has survived 8 moves so far.  It spent several years as the office desk for a local ophthalmologist. I retrieved it when that office closed.  

It is 24" deep, 29" tall, and 42" wide.  

In Use. TI994A workstation.

One flaw in the original design is that if the front of the desk is lifted the desk may collapse. This "feature" is a source of surprise and consternation for the unsuspecting. Curious spectators are tempted to lift up the front.  That challenge was solved with a strap that attached to the center of the top stretcher and connected to a slot in the underside of the desktop.  The strap is strategically designed to also let the desk fold up. 

My award winning success with the PhDesk, Fit Lounge, and an unfolding Z chair design caught the attention of the university and led to them hiring me to teach the furniture design courses. 

Comments are encouraged!

4D    

  

Leaning Shelves

These shelves were a quick and dirty project I made sometime before 1988.  With a wobble dado blade on my radial arm saw it was easy to cut slots for all the shelves on the side posts. I cut the slots before rounding over the edges of the posts to reduce any blowout consequences of the dado cuts.

Lean Too?

The shelves are 2' wide and 10" deep. The system is 68" tall. Shelves are 18mm thick Baltic Birch plywood. Two screws through each post side hold each shelf in the dado slots.  The screws are countersunk with my intention to eventually cover them with plugs.  At the time I made these shelves I didn't have a plug cutter so that task was put off and obviously forgotten.  They have been in constant use since being made. My current house has several 2' wide wall sections that these shelves have leaned on over the time I've lived here. 

The advantage of leaning shelves with no back legs is that they don't have to deal with tack strips under carpet edges that make conventional shelves tip forward slightly. When the posts lean against a wall the shelves have a 1/2" gap from the wall.  Great for power cords if needed. 

For Scale

What is kept on the shelves has changed over the years though.  The TI994A PC and cassette player with a tape carousel was my first personal PC.  I won it in a drawing from a Target store opening day celebration in 1980. Learning to program it changed the course of my life.  Programs were saved through a modem to audio cassette tapes. A floppy disk file system came later. I still have this computer and all the accessories/programs. It hides in a box that is hidden somewhere in my house. 

Comments encouraged!

4D

Parsons Table. A Simple Plywood Table

One of the most frequent inspirations I have for designing and building a project is the presence of scraps of wood left over from a previous project. I use a lot of Baltic birch plywood, and had several roughly 2" wide strips of 12mm plywood left over from a cabinet project.  Those scraps were begging to be made into something useful. 

I appreciate the simple beauty of the classic parson's table. Creating one from wood usually requires careful mitering of pieces and challenging clamping for assembling. There is also an inherent weakness in simple miter jointing.  Strength requires the reinforcement of a splined joint, biscuits , or a locking miter cut done with a special router bit. All of that, to me, is too much work for left over scraps of wood.  A simpler approach was necessary. 

Simple

My plywood table.  Butt-jointed left over plywood scraps, this Parsons table is the end result of my simple approach. 

The strength of this table lies in the interlocking of pieces at each corner of the table.  The beauty lies in how beveling the corners balances each face and eliminates the distraction of the plywood's joint line.  It celebrates the plywood's edge. 

Beveled Outside Corners

This table is 18" square and 20" tall.  The 20" square glass top finishes it off. The green tint of the glass compliments the green color of the base. Four small silicon buttons keep the glass from sliding around. 

Sixteen strips of 12mm plywood were required to make this table.   Eight were 1.53" (38.8mm) wide.  The other eight were 2" (50.8mm) wide.   Glued together the legs and top frame are 2" x 2".  

Top Plane

Top Frame

Legs

I cut a 1/8" deep recess in the top for a piece of 1/8" thick glass to sit flush with the frame top. Before getting a piece cut to fit I realized I had a 20" x 20" piece of glass on hand. The table looked better with that glass sitting atop the base so it won that job. 

This table ended up being a great use for some plywood strips. The top corners have a void and the light colored filler is a 3D printed piece of wood filled PLA, Once I find (or buy some more of) the paint I used for the base I'll touch up the corners to match.   

Questions and comments are encouraged.

4D 

Desk and PC Cabinet. An Ancient Project

 In the years before flat LCD monitors came to be we had boxy analog CRT monitors to view the content of our computers with.  Over the years I designed and made several desks and cabinets to hold the PCs and Monitors and Printers used. Most had some cable management built in. The cables needed then are still the cables needed now and have always been the eyesore that came with PC use. 

When I was in grad school in 1987 I designed this desk/cabinet for myself. It was a challenge to hide away the aesthetically distracting PC parts but make them available when needed.  

Antique Computer Cabinet

This photo is the only surviving evidence of this project. The project was on display at the 1988 IWF Design Emphasis competition where it won first place in the graduate category. The drawer with the PC front showing also holds the keyboard that can be pulled out and placed on the desktop.  The monitor can close up flush with the cabinet face when not needed. A hinging up top door and another door from the side can close down to hide the dot-matrix printer. A shelf below the printer held paper to feed into the printer through a slot above.  

Monitor and printer cables are thoughtfully routed through the hollow left post that supports the upper cabinet. They arrive in the rear of the lower cabinet to plug into a power strip or the rear ports of the PC.  There is a rear panel that pops out to access the rear of the PC and all cable connections.

The desk top and cabinet panels are birch veneered plywood. All edging was hard maple. The desktop is a 5' section from one 8' long piece.  The 3' section left became an alternate top that could replace the long top for a more compact installation. Both used snaps on their end to snap to the cabinet posts. The smaller top didn't need a leg support and would simply cantilever off the cabinet edge.  

The leg post under the desk top was turned on the lathe. It started as three 120 degree sections so tapered slots for the feet could be routed out before being glued together to turn. The feet wedge into these slots and support the post a couple inches off the floor. The post top sockets into a lathe-turned flange screwed into the bottom off the table top. The post with feet could be rotated to point any direction The post and feet come apart easily from the top and each other to make it much simpler to transport. 

The bottom of the posts had wheels attached. With the desk top detached the cabinet could be tipped back and wheeled around. The PC parts inside could remain connected while being moved. A single plug to external power was the only visible cable to deal with. 

This cabinet with the smaller top lived at a sister's home for a few years. It held her computer, monitor, and printer.  When she didn't need it anymore I retrieved it. 

The cabinet was taken apart several years back and the pieces have provided material for other projects.  

I still have the large table top. It replaced the cheap plastic table top of a folding table and has become a project table I can set up when needed in my shop. Legs fold up and the table stores standing flat against a wall. 

Questions and comments are encouraged.

4D 

A Folding Lounge Chair

Fit. Red oak,  Cherry Stain

This chair I designed and built in 1981. I call it my Fit Lounge.  Look up the word "fit" for a rounded out description that very well matches the chair's features. 

Dictionary.com: Fit

Fit is a good example of 4th dimension design.  For most of its life it stands still and ready.  Welcoming me in for some time off my feet.  Several times in its life it has folded up flat for easy carry and compact shipping. 

4th Dimension Design.

The cushions are supported by canvas.  Under the seat the canvas is stretched drum tight using two aluminum bar cam levers. Originally the cam levers were made from oak. Not being sure how long the oak levers would last I replaced them with some 1/4" thick aluminum bar stock.  Through the cushion you never "bottom out" and feel a hard surface.  During shipping/moving the cam levers can release the tension to let the canvas relax.  The back canvas wraps around the top rail, runs over two elastic straps, around a 1" dowel to the low base stretcher between the back legs. It runs around that stretcher and with sewn in Velcro strips it sticks to itself. When the chair is folded up the path for the back canvas is a little shorter, easing tension on the back straps.  The back straps add a flexible lumbar contour to the back cushion. Much more comfortable than a flat plane. 

Canvas Support

Pivot points between the seat and the outer frame have a 1/16" thick nylon washer to keep the wood from rubbing against wood.  The black bracket is 1/16" thick aluminum plate sprayed black for contrast.  If I made a new one I would have the plates powder coated. The cushions have been recovered twice, and are due for one more update. Perhaps with leather this time. 

Folded up it is four inches thick with cushions removed.

All outer edges are rounded over.  A friend volunteered to model it for me.

Gretchen

Fit won 1st place in the summer casual furniture category at IWF's Design Emphasis competition in 1982.  I mentioned to the judges that I also had a redwood version that I took out to my deck when the weather was nice.  Because it folded up flat for easy carry it was easy to take in and out.  

Fit has proven its durability and usefulness.  It has been my favorite landing spot after a hard day of work for 41 years. 

Questions and comments are encouraged!

4D

CNC Efficiency: Tenon Cutting

Over the last decade or so I've been cutting joinery on my CNC for college student furniture projects.  Often the job is simple straight end tenons with the part clamped vertically.  It took just a few experiments to find the quickest and cleanest strategy show in the video below.

For a clean shoulder I used a spiral upcut router bit, running clockwise for a climb cut. This way the fibers are sheared off rather than pushed out. 

The 1/4"diameter router bit has 1" of cutting height and in normal profile or pocket cuts uses 1/2 of the diameter (1/8") for each pass.  That leaves 7/8" of  sharp edges that are rarely used.  A little math to calculate the area being cut in a normal profile pass and you get .25" x .125" =  1/32sq.in.  For a stepover pass that uses 40% of the bit width that area is even smaller at  1/80sq.in. So long as the area being cut off remains between those two the depth of cut and stepover width can be changed. 

For a 3/4" tall tenon 2 passes at 3/8" depth (rather than 6 passes at 1/8" depth)  reduces the time spent running around the tenon by 2/3. To keep the cut area under 1/32sq.in the stepover should be no more than 1/3 of the bit width. 1/16" width is a safe amount for a 3/8" deep pass.  Roughly 1/43sq.in. per pass.  The whole 3/4" depth could be cut in one pass if the stepover was only 1/32". In that case the chips produced tend to be 3/4" long fibers and are more challenging for a dust collector. That is the reason I make the pass depth 3/8".   If your CNC has some backlash/play you can run a conventional counter-clockwise final cut around the tenon to trim off any not cut by the climb cut.    

Another variable is feed speed.  The short lengths of the tenon sides self limits the speed of the bit travel. Acceleration/deceleration between nodes never reaches the feed speed set. This is true for the small Probotix CNC I use. For beefier and quicker CNCs I'd set feed speed to 100ipm. 

If you are using a larger diameter bit then the stepover and pass depth can be even larger. Do the math to verify you are still getting good chip removal. A 3/8" bit, taking 3/16" deep passes for a profile cut removes 1/14sq.in. as it moves forward. Making a single 3/4" deep pass taking off 1/16" is 1/22sq.in. removed.  Easily managed by the 3/8" bit. 

This strategy also works when the tenon is being cut at any angle up to 20 degrees or so from vertical. For woods that tend to split easily a smaller angle limit is recommended. 

Questions and comments are encouraged!

4D    

Knock Down Panel Joinery: Tusked Mortise and Tenon

Tusked Mortise and Tenon

This is another CNC cut panel connection, It is a variation of the more classic Through-tenon with a wedge pin often used on trestle bases of dining tables or workbenches. I made this sample to show my Workshop 2 Furniture Design students as one option for their knock-down furniture design assignment.  

The Wedge and the Slot 

Both the wedge and the slot for it were cut out on my CNC.  The inside face of the slot is sloped to match the slope of the wedge.  The fluting toolpath in Vectric's Aspire or V-Carve software made easy work of the slot's inner slope. 

Tenon and shoulder

The shoulder of the tenon has a dogbone cut to account for the inside corner which a round router bit can't make. I added a 0.003 allowance to the slot for the tenon for a nice slip fit. The slot for the wedge has room for the wedge to pull the tenon tight against the back of the joint.  Sloping that slot to match the angle of the wedge face makes a high friction interface that stays very snug with the wedge tapped in. A tap or two on the bottom of the wedge will pop it free to make disassembling the joint quick and easy.  

Comments and questions are encouraged!

4D

Knock Down Panel Joinery: The Twist Tenon

Twist to Lock Tenon

For the Workshop 2 furniture design classes I taught I had the students design and make furniture that could quickly assemble and also knock down (come apart) easily. 

Most woodworking joinery is not designed to come apart repeatedly. This CNC cut twist tenon connection does slip together easily and once twisted 90 degrees locks the parts together securely.  Twist the other way to allow pulling the parts apart. 

Twist 90 degrees to align with the center slot

This example is a snug fit with only a few thousandths clearance between tenon and the slot it slips through. When twisted the fit is still snug, but could be improved by adding a slight ramp to the twist surface on the end panels. That would have the twist action wedge tighter as it turned.  The fluting toolpath in Vectric's Aspire or V-Carve software could handle the ramping. 

Once apart all pieces store flat.

Using this joint a simple table could be made from 4 panels. The center panel would have tenons on two sides and the top. Rotate the sides onto the side tenons, then rotate the top onto the top tenon.  With the top in place it would prevent the sides from rotating. A double thick top could have the top tenon recessed flush when assembled. A couple of spring loaded pins could lock the top to the sides as it rotated into position. 

Comments and question are encourage!

4D