100 False Ribs Done

While I've been writing all this I've also been making ribs. After 100 ribs you get very good at it. That many ribs makes quite a stack of ribs.

Now that they're done I did one more experiment. I wanted to know if this would work with 2024-T3 aluminum instead of the 5052-H32 we first selected. The 2024-T3 is much stronger but also has more spring back and I didn't think we could form the channel walls square. The 2 ribs at the top are 2024 and the rib at the bottom is 5052. Clearly the form block needs more curve to allow for the spring back of 2024 but otherwise it works fine. It does take much more force to make the ribs. The channel on the right is 2024, again no problem with the shape. Also there is no evidence of cracking along the corner radius even looking with a 10x glass.

I'll give some thought to applying the same design process to making a form block for 2024-T3. I also want to do some load testing to see if there is any real advantage to using the stronger aluminum.

Forming The Ribs

The process of forming the aluminum strips into ribs is fairly simple. You do need the fixture solidly clamped down.

With the roller down at the nose of the rib loosen the clamp and turn it to the side. Insert the tip of the strip into the notch and bend it back flat to the form block.

Position the clamp and tighten it down.

.

.

.

.

.

.

.

.

.

Brush oil on the strip and the roller. I'm using 5w30 motor oil but there are probably better oils to use.

I made wooden blocks to hold the strip from moving side to side. They have a notch band sawed into them to slide over the strip. The cut is deeper at the front than the back. Also the wood below the cut was sanded to curve the surface so it can wiggle as needed to clamp the strip centered on the form block.

The sides of the blocks are oiled. The blocks are clamped tight with a plastic clamp from Lowes. It flexes to work like a tight spring.

.

.

.

.

.

.
.
.
.

The forming starts by lifting the handle and pressing the roller down into the strip while rolling toward the Trailing Edge.
After a few inches stop and roll back over the start to tighten it up. Then add a small plastic clamp to the end of the channel to keep it from lifting from the form block.
.
.
.
.
.
.
.
.
.

After about 6 inches the roller stops. I believe the angle is off slightly. By holding the handle down and pushing it rolls on just fine. It doesn't form the channel as tight this way so you need to tip it back to the normal position and re-roll it. I also could have used a slightly longer form because with it tipped down it misses the last bit of the rib.

.

.

.

.

.

.

.

.

.

This last bit of the rib is finished when you re-roll with the handle in the up position so it's not a big deal. I've taken some pictures so when I get some time I'll see if the stopping problem can be fixed.

At this point the rib is formed. Wipe off the oil, remove the clamps. The roller needs to be wiped off before moving it back to the start. It picks up a lot of aluminum during the rolling. Cleaning it prevents galling during the rolling of the next part.

The channel is formed with a nice radius in each corner.
The bent end gets cut off leaving a 1/2" tab for nailing to the leading edge and you have a rib. Just 99 left to make. A few more hours work but very satisfying.

If you didn't see the video of this in the earlier post click here.

362 Rib Form Block

The complicated part of this was not making the parts. First I needed to take what we learned from the first tools and us it to shape the tool so the parts came out to the correct shape in one rolling operation.
To do that I traced the tool and the formed part so I could re-draw them in the CAD system. I needed to look at the bend angle created in the part vs the angle in that segment of the form block. To do that I divided each curve into 3/8" segments and measured the change in angle from segment to segment on the block and the part. Then I plotted this data to get the relationship. I tried 1/4" segments first but the data just had to much noise to use it. Now all I had to do was take the WACO part drawing and divide the curve into 3/8" segments and measure the angle between each. By applying the relationship in the plotted data I could now draw the curve for the tool to form .025" 5052-H32 aluminum. OK, I could have just done this with calculus by developing the equation for each curve and calculating the first derivative, etc., etc. I'll leave that for someone who has used their calculus skills more recently than I have. I'll save that for a retirement project.
With our new drawing printed out to full scale it's just a matter of gluing it to the aluminum and rough cutting it out on the band saw. The shape was finished to the line on the belt sander and then hand filed smooth. The 1" grid lines are there to check that it really printed correctly to scale. Holes were then drilled to allow screwing this piece to some angle for clamping it in the vice. John milled a nice curved hole in his, I used the band saw so I needed a way to hold the curved strip of aluminum while forming the parts.

The holes were tapped to allow screws to be threaded into the holes.

Another scrap of 1" x 1/8" aluminum from our old Ultralight project and we have a nice base for clamping in the vice.

When I made the roller I left as much metal on the inner roller as I could. As a result it was easier to use the actual parts to determine the spacing between the rollers to get the desired angle. This was done by laying the parts over the drawing and measuring the distance. Clearly I could have done this in CAD but it took less time this way than booting the software.

All the parts to assemble the roller handle.

.

.

.

.

.

.

.

.
.

The roller handle and form block assembled in the vice.
.

.

.

.

.

.

.

.

Next we needed a clamp to hold the end of the aluminum strip. I had band sawed a slot into the form to hook the end of the strip, but the roller will pull the strip out of it without a clamp.
The first clamp I made was to long to turn side ways with the roller down, and too tall for the roller to pass over it to the start position. The shorter clamp works perfect. It has a heel on it to allow for the thickness of the aluminum strip.

.

.

.

.

.

.

.

.

.

The last change was to shorten the tip of the form block so the roller can move down far enough to make it easy to load the strip and clamp it. Also the area around the screw socket had to be filed and polished to stop it putting marks on the aluminum roller.

We're almost ready to form parts.

Adjustable Form Roller Design

I had 3 variables I wanted to be able to adjust in my experiments with the Forming Roller. The first was the width of the gap for forming the walls of the channel. My belief was that making that gap exactly 3/8" + 2 times the .025" aluminum, i.e. a line to line fit would be to tight but how lose should it be. I could get John to turn it line to line and than keep re-turning the gap wider but it is very hard to get the roller re-chucked square and true. Also once you take off to much you can't put it back. The result was I wanted a roller made in 3 pieces, the outer walls and the inner roller. Then I could make it line to line and shim the walls out to add gap until I found what worked best.
The second thing I wanted to do was change the depth of the channel pocket to allow more wiping of the sides of the channel to assure square flat walls. This one I only needed to take down until it worked. A little over shoot on this was not likely to be a problem but it still meant re-chucking the roller.
The 3rd variable was the angle of the lead on the the roller walls. The 45 degree walls seemed to work but I wanted to be able to make them steeper if needed.
Oh yeah, John could and would be happy to turn the roller but I tend to work on this stuff in the evenings (after 12 hours in the shop) so I really wanted to make the parts my self. John lives about 30 miles from his shop and has a young family to occupy his evenings. I also had never gotten my lathe set up after building this house so all this had to be done from flat bar stock using a band saw to cut out circles, a belt sander to smooth and shape them and a drill press for the needed screw and bearing holes. Oh yeah I really wanted it all within a few thousandth of an inch square, smooth, round and concentric. What fun is life without a wish list. I bought some 3/16" x 3" wide steel from Lowes, some 1-1/8" diameter roller bearings from Tractor Supply, and a cheap hole saw, screws and a new tap from the hardware store. There was enough scrap from cutting the form from the 4" wide 3/8" aluminum to make the center roller from it. I probably would have liked a steel roller but that meant ordering bar stock from McMaster-Carr. If aluminum didn't work that would be my back up plan, you gotta have a plan.
The first step was to saw the 1-1/8 holes in the steel and the aluminum. There would be no way to hold the disk once it was cut out or to get it very square to the hole saw. Oh yeah did I mention I had to buy enough steel for about a dozen disks. I was just delighted it would work. The edges aren't square but we're putting a big chamfer on them anyway. I drew the circles the full 3" width of the metal and just don't worry about the little flat spots on 2 sides of the disks.
.
.
.
.
.
.
.
.
.
To make the outside diameter concentric with the bearing hole I bolted a stack of 3 washers (AN970-4) that just happen to be 1-1/8" in diameter to a scrap of plywood. There is a little clearance between the hole and the washers so by clamping just one end of the wood to the belt sander table you can wiggle the other end closer to the belt. By pushing the disk against the washers while rotating it you can slowly sand it round while keeping the O.D concentric with the I.D. It gets hot so I would do one a little then do the other. Eventually all the bumps are worked out and the the 2 parts are round, concentric and the same diameter. You don't move the wood in at each step until no metal is being removed as you rotate the disk.
The same process was then used for the inner aluminum roller.

By tipping the table to 45 degrees and using the same process the edge of the outer walls we chamfered while keeping the chamfer concentric as well.

The surface finish is rough but the shape is spot on.

.

.

.

.

.

.

.

.
.

A test fit and it looks like our roller. So now the ground surfaces need to be smoothed without losing our shape. Break out the 10" mill file. This works if you can turn the parts so we need a way to spin them. I had a rubber mandrel for a 1" sanding drum. To small to fit our 1-1/8" holes. I also had a rubber stopper on some 5 gallon water jugs I use in the field for measuring the permeability of soil (how fast water flows through the dirt). I had already drilled a 5/16 hole in the middle which fit the mandrel perfect and they were about 1" tall. So now I just needed to cut and sand one down to 1-1/8" just like I did the disks. Hey it worked on steel and aluminum why not rubber. We used to grind rubber parts at AC by freezing them so I knew it would work if I was careful not to overheat the rubber.

.

.

.

.

.

.

.

.

.

Voila! A mandrel to hold the parts in the drill press while lightly filing them smooth.

I know it looks nutty but it works and they came out great. Also all of this works if you need to change angles on the chamfers or the diameter of the inner roller without losing the concentricity of the parts. It was easier to file the outer disks with the chamfer turned out.

Now we just need to screw all this together and we'll have a roller. By using the mandrel to hold the parts concentric, and my father-in-law's machinist clamps to hold the stack tight, holes can be drilled through the stack. The holes have to be the diameter for tapping the holes. There are six holes. Three flat head screws from each side are tapped into the opposite side. Once the tap holes were drilled the parts were disassembled, marked for re-alignment and the clearance holes drilled & Countersunk.

.

.

.

.

.

.

.

.

The six hole were then tapped. One thing I would do different is to use the center roller as a guide to keep the tap square. One hole was off just a little from square making the screw a little tight. It worked but the problem could have been avoided.

The screws were a little long so I used a washer on each one as a guide to belt sand the screws to length. At this point there are no shims to make the clearance for the .025" aluminum so the washers worked perfect. As soon as the belt kisses all three you're done.

Shims were made from .020", .025", & .032" aluminum sheet. With a .025" shim on each side it's line to line. I started with .020" on one side and .032" on the other which give .002" clearance. I tried other combinations but that worked perfect, go figure.

.

.

.

.

.

.

.

.

The finished roller. The only other thing I did was to polish the working surfaces with crocus cloths, and then jewelers rouge on a pad with the Dremel tool while spinning the disks on the mandrel, nice finish.

On to the form block.