You will need a trade supplier of secondhand industrial plastic drums, and the Yellow Pages is the best place to look. This will be a 'drum laundry', where steel and plastic drums for industry are washed out, then sold clean, for material suppliers who find new containers too dear. You need a blue high density polythene drum, of 150 litres capacity, with flat sides and a small black polythene lid. A washed-out drum will still smell of its previous contents. Industrial soaps, detergent paste, or surfactants are the best, since all are virtually non-toxic. Phenolic resins, which stink, are best avoided. Drums that hold phenolic are rare, and are hard to resell, so your supplier may be tempted to offload one on you. Drums that have held dyes are not too bad, unless strongly stained inside and out. The drum should not be dented or deeply-scored. Make sure the gasket inside the lid rim is in good condition, along with the top rim of the drum. The lid should fit snugly, and seal down with moderate pressure on the handle of the sealing ring. If the lid or ring needs to be thumped with a mallet, it will be no good out in the snow. Try another lid and ring, and offer the supplier more for a good one if you have to. Secondhand drums are cheap, and it is worth paying considerably more than trade price for a really nice example.
The drum laundry only does a quick job of cleaning out. Be prepared to spend most of the afternoon scrubbing the drum out with methylated spirits (industrial alcohol), detergent, and water, then a mix of all three. Most of the smell of the previous contents will have gone when you finish, and the remaining whiff will fade away over a year or so. When you have a sparkling drum with a well-fitting lid, you can either carry on, or use it to store dry sleeping-bags in dirty sheds.
Mounting the wheel, and bearings, on a shaft is not straightforward. If you imitate the trolley makers, and just push the bearing assembly onto a steel rod, with a split pin and washer, you take pot-luck as to whether the ball-bearings spin as they should, or whether the bearing sleeves rotate on the rod, increasing tolerances, and wearing fast. The wheel is going to take a hammering on the sledge, and bearing mounting like this is not adequate.
If you cannot turn up a short axle yourself, get a turner to make one for you. I know the bearing sets are only pressed into a plastic wheel, but this makes rigid mounting of the sets on the axle even more important.
Changing units quickly, my axle is 71 millimetres total length. There are 25.4 mm to the inch. The steel axle has a short 12 mm wide section, where it mounts on the aluminium stub, and this section is 40 mm in diameter. A long hole is drilled down the centre of the axle from this end, and tapped 8 mm metric, or the equivalent in Whitworth or UNC. Provided the aluminium stub is sturdy enough, a length of stainless 8 mm allthread can take enough torque to firmly locate the 40 mm diameter mounting surface.
The photo below shows the axle held by the allthread to the end of the aluminium stub. Note the axle terminates flush with the stub, and an aluminium turned cylinder of the right dimensions (see later) has been forced into the stub to provide a rigid end. The other end of the stub can be seen firmly welded to a piece of plate, which is held to the reinforcing on the base of the drum by four studs. In the free end of the axle is an 8 mm socket head screw, which tightens up to grip the wheel bearings. The turned aluminium sleeve inside the wheel, between the two wheel bearings, is not shown.
The hardest part of the project is now over. The wheel will not feel perfect, as it rotates on the axle, since the bearing sets are not industrial, but there should be only be a small amount of free play. The odd 'tight point' is also to be expected, and will 'run-in' with use. After five long trips, the wheel on my Mk 2.5 is running perfectly.
Cut off a few inches of 8 mm metric stainless allthread, or equivalent. Lock two nuts on it, and wind it hard into the tapped hole in the axle.
At the very minimum, drill a corresponding hole in the other side of the square stub. Pass the allthread through both, screwing on extra nuts, and adding big washers where appropriate, as you do so. Tighten the washers and axle down as before, then lock a pair of nuts and included washers down on the allthread, where it passes through the other side. This may give you enough strength.
This is lathe work again. Turn a short length of round aluminium bar down, until it just pushes into the end of the hollow extrusion. Part off about 40 mm of this turned bar, put it back in the chuck, and take a little off the end, until it just starts to fit sideways into the hollow extrusion. Using the lathe, drill an 8 mm hole right down the axis of the bar. Push the bar section sideways into the top of the hollow extrusion, and tap it down with a hammer until it is just inside. Since it is an each-way fit, it will align, but you may need to hit it quite hard.
Once the bar section is in position, drill by eye a 5 mm hole through the side of the extrusion, so it ends up inside the 8 mm hole already drilled through the bar. Use a small round file to open up the hole until it matches with the 8 mm one. You can now electric-drill down through the 8 mm hole and through the other side of the hollow extrusion. You now have a solid braced end to the stub. Push the allthread sticking out of the axle through the 8 mm hole. Screw on a single nut and washer, and tighten down. The excess allthread can now be sawn off, and the cut filed smooth. There is now a maximum-strength joint between the aluminium stub and the short axle.
I faced off the end of the extrusion in the lathe, so it was accurately square. I cut with a hacksaw, a 5 inch square piece of 6 mm (1/4 inch) tempered aluminium plate, and drilled an 8 mm hole right in the centre. Two nuts held an odd length of 8 mm allthread firmly in this hole, long enough to poke out of the top end of the stub. Giant specially-made washers provided alignment, at both ends of the stub. I screwed down a single nut, on the top of the stud, and it was firmly mounted in place, ready for welding.
Because accurate clamping is the hard bit, the welder charged me very little to lay down a thick fillet on the four-sided joint. I cleaned up the weld with a coarse round file, drilled four holes accurately square in the corners, not too close to the edge, and rounded the corners. This completed a wheel assembly that normally lives off the drum, and is bolted on when I arrive in the mountains. Despite frequent impacts the plate/wheel assembly remains as straight as when it was made. The tyre, wheel, axle, stub and mounting plate weigh 2 Kilograms total.
I widened the aluminium hole to the self-tapper cutting size, pre-cut the thread with a spare self-tapper, then put the aluminium back on the drum base. I screwed a self-tapper into the aluminium from inside the drum. This located the aluminium correctly at one point on the drum. All I had to do was repeat for a second hole, and I had location.
I located a total of 8 self-tappers, to hold down the reinforcing plate. Inside the drum each one passed through an oversize washer to avoid distorting the polythene. The screws were too long, and I cut off the excess on each, and finished off the rough cut on each with an angle-grinder. I could have sealed each screwhole with Silastic or similar, if I anticipated taking the drum through deep water. The aluminium plate was now mounted adequately on the drum, with minimal effort. After an Autumn and Winter of heavy use, it has not worked loose.
I put some blobs of 'blu-tak' putty-type adhesive on the contact area of the welded plate/aluminium stub assembly. The wheel was firmly bolted on the stub. I tapped the plate down on the reinforcing aluminium, shifting it round until the wheel was central on the end of the drum, and facing at right angles to the snow sliding surface. I then hammered down the plate to get maximum adhesion, and took the wheel off.
I electric-drilled through the four 8 mm holes, only pushing far enough to cut small drilling centres into the plate surface. Everything was then dismantled, and I machine-drilled four true holes through the marked centres before tapping them with an 8 mm thread. Tapping was started in the drill chuck, to keep the threads precisely at right angles to the surface.
I then put the reinforcement back on the drum, and used a smaller drill to run though the four tapped holes into the drum, without damaging the threads. Off came the aluminium once more, and I oversized the four new drum holes to ample clearance for 8 mm socket head screws. Back went the aluminium, all the self-tappers were tightened, and four socket head screws went from inside the drum, through pancake washers, and were moderately tightened into the aluminium tapped holes. This further located the reinforcing, as well as locking the screws, so the protruding studs were both rigid, and resistant to rotation. Work on the bottom of the drum was completed.
The protruding studs were long enough to drop the wheel mounting plate over them. I tightened four nuts and washers down on the plate, and the wheel was firmly mounted on the drum. The whole wheel assembly is normally stored inside the drum, and only mounted when the car trip is over, and the hiking/skiing begins.
Because the reinforcement is inside the lid, resistance to rain and wet snow is not so important. I used half-inch thick exterior grade plywood, which is cheap. A circle just over 12 inches in diameter fits neatly inside an internal ridge on the lid. The lump of ply I had handy was just too narrow, so I have a couple of small flats on the circle. I originally used 8 fat self-tappers running into the ply spaced equally around the ply rim to provide edge mounting, but four of these became redundant. They were replaced by locating points for the combined ski rack and towing skirt holder.
The photo shows many other screwheads running through the top, but this lid has been used for umpteen variants over the years, and most of these just plug holes to keep the water out. A few self-tappers are a good idea closer to the centre, just to stop the inside of the lid from flapping up and down.
The 'house' is 50 mm wide, and 46 mm deep. Each eave is 15 mm above the base. The top roof line is 39 mm above the base. The shallow roof is 36 mm by 50 mm, and accurately flat. The steep roof is 26 mm by 50 mm. Precise dimensions are not essential, as long as the block is true, and not skewed, but don't make it much smaller than this. It might be possible to use high density polythene, or polyacetal, to save weight, but I would not take the risk.
Allow some waste at each end, unless you are being clever, and scribe a line across the block at an angle of 54 degrees. This corresponds to the 36 degree slope on the shallow roof. Square the ends of the scribed line across the block carefully, to the other side, and scribe a corresponding line, across the opposite face of the block. Unless you have your front and back lines crossed, you have now defined all four sides of an angled cut. Slowly work round the cut with your hacksaw, taking about a millimetre out before shifting to the next side. Just go round and round. If your initial scribing is true, you will make a true cut. As the cut deepens, you can take more out, and the whole process accelerates. After a good while, you will have two halves, and each half makes a 'house'.
Rub a thin layer of engineers' bearing blue on a piece of flat glass or similar. Rub one of the cut faces of the block on the glass, and blue will transfer to the high points of the face. File down the blued high points a fraction, and do another rub. Repeat until the cut face is flat enough for you. For engineers reading this, and going purple, yes, this is the incorrect method of using bearing blue, but it is good enough for this job. If all is going well, you will be gawping at two machine-flat faces, and wondering how you did it.
Each 'house' is 50 mm or 51 mm wide, depending on whether the block was 50 mm or 2 inch. As well as the steep roof, both of the 'house' sides are flat, since they are original block sides.
The cutting of the two mounting blocks is finished, and they will look good. Attention to detail is not only cosmetic. Unless the faces, especially the shallow roof, are flat and true, the handles will not work properly.
The photo shows the plug after fitting in the end of a handle tube, with the locking hole drilled, and a short small self-tapper screwed in each side
Take a faced-off handle end, and run a Stanley knife blade, or the business end of a large new screwdriver, around the inside of the end of the tube, to take off any burrs. You will overshoot when fitting if this is not done carefully.
Reduce the diameter of the end 42 mm of the bar. Try the deburred handle tube on it frequently. The object is to get the bar down to a size that will drive up the inside of the handle tube, with no slack. Once you can push and wiggle the handle up about a quarter of the length of the bar, the diameter is right. Wiggle it slowly off again. Part the plug off from the chuck, leaving a 36.5 mm section about half an inch wide. This is ample for rigidity. Repeat this sequence to make a second plug for the other handle. Drawn tube fluctuates in internal diameter, and your two plugs should be individually tailored for the two handles.
Mount a plug in the three jaw chuck, with the 36.5 mm shoulder against the jaws. Drill out a tapping size for 8 mm metric or equivalent. Again tap alignment is critical, so get as much as possible of the thread cut while holding the tap in the lathe drill chuck. Finish the tapping with a normal tap wrench. Go in as far as you can. There will still be partially uncut thread, for the stud to bite into.
Support the other end of the handle tube on wood, and drive a plug right up to the wide section. Repeat for the other handle. Despite the tight fit, repetitive flexing will work the plugs out during use, so they need to be locked. I could have used epoxy, or even super-glue. Both presented assembly problems, as well as being impossible to get apart again. I put the plugged tube into the vice, and with an electric drill, put a hole about 10 mm back from the handle/plug joint, running right through the tube and plug, until it came out the other side. Luck was with me, and it was near enough central and perpendicular to fool the eye. I forced a short slim self-tapper into each end of the hole, solidly locking the tube and the plug together. The self-tappers were too short to poke into the 8 mm internal thread, so it remained clear. I repeated the operation to mount and secure the second plug.
With the plugs locked in the ends of the handles, hold this end of a handle in a vice. Cut off a few inches of 8 mm stainless allthread and lock two nuts on it, Wind the allthread into the plug until the thread seizes. Hacksaw off the allthread, so there is about 17 mm poking out. It is worthwhile slipping the handle back in the lathe chuck, and facing the end of the allthread off perfectly, so it engages easily with the thread in the 'house'. Repeat for the second handle and plug. Apart from final length adjustment, your screw-in handles are completed.
The towing skirt has been a standard fixture in many of my oversnow designs. It derives from the large bearskin, or similar, that was used to pull dead animals, or bundles of firewood, over deep snow without the need of a sledge. The front of the bearskin is tensioned by the pull, and progressively packs the snow as it slides forward, so the load slides along a packed strip of snow rather than sinking into it. Almost any short fibreglass or polythene pulk will give trouble in soft snow, jerkily sinking into the snow, so not even the aluminium towing tubes will smooth things out. When the snow is fresh and sticky, such designs will push great lumps of snow in front of them as well. Even if you make nothing more of the sledge, using a towing skirt, ahead of existing pulks, sorts out a great deal of trouble. If your pulk glides rather than bucks along, rope towing becomes much more acceptable, and there is a good deal less to stow in, or on, the car.
The material used for the skirt is clear low density polythene sheet, from an industrial plastics supplier. Thickness is one sixteenth of an inch, or 1.5 mm. Width of the skirt is 16 inches, and this figure is quite critical. The total length of the polythene used is 34 inches. If you shorten it more than 6 inches, and tow through deep light powder, you will be in trouble. Snow that flows round the legs just below the knees, while just the ski tips are visible, is a great rarity in Australia. Basically, all you need is two 16 inch lengths of wood, one tacked to each end of the strip. One bolts to the sledge, and two 33 inch towropes go from the other to the skier's waist.
In practice, things are more sophisticated. Wood will get wet, and deteriorate, so the strip that contacts the sledge is made from high density polythene. The strip is 1.5 inches wide, and 1 inch thick. I had antistastic black grade available, pre-cut to a strip, so used it. Again, an industrial plastics supplier will saw it for you. If you really do like complexity, a light drawn aluminium tube 40 mm diameter would save weight, but create construction problems.
The towrope end can be much lighter, and extra weight here can upset the whole tow. I did not want to use wood, so sandwiched the end of the skirt beween two lengths of half-inch square hollow aluminium extrusion, 1.6 mm thick. Close-spaced long 3 mm metric screws ran from one extrusion, through the skirt end, and into the other extrusion. This has been 'make and forget' and the whole assembly has given no trouble.
The towropes are short at 33 inches, since all of the skirt is towing length. They finish in loops. Strongly hand-sewn on the padded hip belt of my pack, are two loops of flat nylon webbing, the same width as car seat-belts. I fit two short cords to these loops, and the other ends of the cords terminate in two slim screw-up steel oval rings. These assemblies live on my pack all Winter. I just drop a towrope loop through the oval ring on each side, and screw them up, making a perfect joint in seconds. The sewn points on the waistband take the force, and towing is comfortable.
Any time a towing skirt is fixed to a lower strip and dragged along the snow, scraping of the heads of the fixings against ice, fallen timber, and the odd lump of rock, has to be taken into account. I mounted the strip vertically on the drum lid, that is at right angles to the towing skirt. I rounded off one of the four hard edges on the strip (a wide sharp wood chisel is best) so that the skirt was fastened well clear of the snow, then wrapped the skirt 90 degrees round the radiused corner, before it contacted the snow. This had many advantages - the slight spring of the skirt on the bent corner allowed it to ride over rock tips etc without gouging.
The support arms were made from 20 mm square hollow aluminium extrusion, 3 mm thickness. Holes were drilled at the right separation for bolting down to the spacer nuts, and I did the normal 'overkill' by turning up cylindrical spacers to jam down the inside of the square extrusion. Look at the wheel mounting on the aluminium stub for details of a similar process. This made sure of maximum strength at the mounting points, and eliminated any risk of the extrusion buckling under excessive tightening. The arms were bolted down onto the 25 mm spacer nuts before cutting to length.
The arms now set the 260 mm spacing for the skirt strip mounting. Arm length is critical, since that sets the tilt angle of the drum relative to the compressed snow generated by the skirt. In practice an arm length, such that a line joining the two mounting points just grazes the outer lip of the lid, is ideal. I drilled 8 mm clearance holes in the front sides of the arms only, since stresses are minimal at this point. I could then cut and smooth the arms to length. Into each hole went a length of 8 mm stainless allthread, locked by nuts internal and external to the extrusion. I measured the precise distance between these two studs, then drilled holes in the strip mounted on the skirt to correspond. I dropped the strip onto the holding studs, tightened up a nut and washer on each to hold it down, and cut off the excess stud length. This completes the towing skirt mount.
I extended the studs holding the high density polythene strip to the angle, and with 25 mm long spacers parted from small aluminium tube, and washers, made a 'spool' at each end of each ski-rack. On one spool each side dropped a 41 inch length of 25 mm nylon braid, held by a sewn loop. Velcro type hook and loop, in just the right place on each side of the strip, made a ski-holding strap that could be wound round the skis and poles a few times, to grip them firmly. Did it work - I'll say!. The captive straps really have made using the whole design a positive pleasure.