February 9th, 2020

This month's meeting was held at Lee Valley Tools.  The theme was Practical Ideas for Working With Wood with a presentation and demonstration by Phil M.

drum_sander_-_phil_main_-_compressed.pdf
File Size:	698 kb
File Type:	pdf

Download File

---

Phil showed us the thickness drum sander he designed and built, that was featured in the Canadian Woodworking Magazine a number of years ago.   A downloadable PDF copy is available above.  NOTE: the magazine artist took some liberties with his design and so their image does not fully agree with our accompanying photos.

He uses it to sand his planks and other pieces to the proper thickness for his builds.  It features a tilting base with a piano hinge on the feed side and tilt adjustment via a fully threaded carriage bolt through a blind nut on the other.  The drum has Velcro face to which Velcro backed sandpaper to adheres (available in a variety of coarseness/grades).  The drum is mounted with two pillow block bearings and is driven by a Vee belt connected to a 1/4 HP 1800 RPM 120 V 60 Hz electric motor suspended under the table.

The dust cover is a separate piece that sits over the sanding drum.  It has see through Plexiglas faces and a vacuum connecting spout on the top.  He admits it is very good at making dust!

He says the wood pieces must be held securely when feeding through and caught up on the outfeed side before letting go on the infeed side or it will throw out back at you.  It is a good idea to stand to one side.  He explained that if he was to do this again he would make the table shorter so he could get a better grip on shorter pieces.  He also explained that the Velcro does make the sanding surface a bit soft so the wood must be fed through at a slow, consistent pace.  Stopping the feed results in an unwanted divot.  The use of calipers are a must to measure sanded thickness results in the work and make table height adjustments for additional passes.

              Bending by clamping to a form

                         Bending with heat

Phil then discussed how he bends wood.  In both cases he places the piece in boiling water (in an old electric kettle) to make the wood cellulose fibres more pliable.  The secret to bending wood is not the water, but temperature.  Cellulose in the wood softens at about 200°F which is near the boiling point of water. The water is simply the vehicle used to increase the temperature.  Once the wood has been shaped, it’s imperative to fasten the piece to the model with masking tape, to let the wood cool and dry. As it dries and cools, it will shrink and the cellulose will hold its new shape.  The normal rule of thumb for soaking is about 1 hour per inch of thickness, so a 1/8" thick piece should soak for 1/8 of an hour or about 7-1/2 minutes.

Phil uses a soldering iron with a flat nib passing it quickly over the surface of the wood (steaming the water in the wood) while he bends it by gently pulling up on one end.

He also does it by clamping the part to a mould or form.  He uses multiple clamps and leaves the part overnight to dry completely in the new shape.

There is a limit to how much a part can be bent.  We also discussed laminating thin strips of wood with white Weld Bond glue and bending immediately, clamping to a form, and letting it dry completely before removing it.  The Weld Bond adhesive dries clear and if not applied too thickly will be completely invisible.

Phil then described how he prefers to avoid painting his models where possible by choosing a variety of lumber that has the natural colour the model needs.  He brought quite a variety of samples to show us, Swiss Pear, Holly, Bloodwood, Purple Heart, Amarillo, Poplar, Basswood (Linden), and Boxwood (Castello).  He explained how he had some Holly that had aged (darkened) so he tried bleaching it with a diluted solution and it came back to it's natural bright white.  Some woods are less suitable for models than others, some will keep an edge better than others, and some can be quite toxic, so you should know your wood and wearing a dust mask while cutting and sanding.

Phil also showed how he uses rubber cement to adhere his paper patterns to wood for cutting.  He brushes it on the wood and lays the paper over it, rubbing it down.  Only applying it to the wood makes it easier to remove the paper.  When done he peels the paper off, rubs the glue off with a crepe eraser  (the type used to clean sand paper) and then scrapes a fine layer off the top of the wood with a sharp blade to remove any glue residue.

He also discussed his preference for Viking band saw blades.  He finds them to be more flexible and last longer.  He finds them to be more flexible and last longer as they are a bi-metal composition.  Because they are installed at a low tension, there is less wear on the tool and they ease the problem of wood moving away from the fence as it does when cutting thin stock with steel blades. These blades are made in Ontario and are available from Lee Valley.

Phil had brought in his 16 gun model (under construction) of the USS Lexington (1776) a two masted Brig which can be seen in the photos above.  He has done a fine job on her to date.    The model is being scratch built following plans by Clayton Feldman and represent an amalgam of a type of small, armed cutters, sailing on the Great Lakes circa 1800.

Ray had brought in his model of the Isabel, a North Sea fishing trawler that had been converted into a personnel sail boat.

He had seen the vessel while on vacation and took a few photos.  Knowing the length after speaking to the owners he then scribed lines on the photo to convert measurements so he might build her.  The hull was shaped from a solid block of wood, as was the mounting base.

Alan brought in a sample of two methods for mounting his keel to his build board and eventually to a display board.

One method used is a simple #6-32 hex nut embedded in the top of the rising wood as shown in the fore portion of the photo to the above left.   The nut is a thick as the rising wood which is sketched in pencil at the end of the block.  In Alan's mind, there isn't much top wood left to glue the frames to.  The outside flats of the hex nut could be filed narrower to reduce the size of the pocket as he had seen volume 1 of 4 in the The Fully Framed Model, but he devised a different approach as seen at the back end of the photo.

Alan purchased some #6-32 Helicoil threaded inserts, 304 Stainless Steel,  and a length of #6-32  zinc plated threaded rod. The dissimilar metals will avoid any degree of galling over time.  He drilled a #29 (0.136" diameter) clearance hole through the rising wood/keel/false keel assembly for the threaded rod (0.132" diameter) .  He then drilled a #15 (0.180" diameter) hole down through the rising wood for a depth of 1/4" for the Helicoil insert (0.1785" outside diameter).  The threaded rod was passed through the hole, thread coated with petroleum jelly (Vaseline) and the helicoil insert threaded onto it.  A two part epoxy adhesive was applied to the outside threads of the insert.  The Vaseline kept the glue from leeching through the thread and gluing the insert to the rod.  The rod was pulled to draw the insert into the stepped hole.  Then the rod was removed and the insert left until the adhesive dried and cured.  The tab was broken off the insert the next day, but actually could have been removed prior to the assembly. 
​
This method leaves a tiny hole at the top of the rising wood, and more wood to glue frames to.  Also worth mentioning is that the hex nut has 3.4 threads where as the inserts have 7 full threads of holding power.  As a mechanical designer, Alan appreciates that it takes 5 full threads to develop the full holding strength of a mechanical threaded assembly.

As an after thought, the pocket could be drilled slightly deeper so the exposed hole on the top face of the rising wood could be closed with a plug of wood providing a full glue surface for frames.