Monday, April 21, 2008

Sunday ride

So I did my first longish ride on this bike. Check out the route. Overall things went well. Total distance 52 miles, average speed ended up being a pathetic 13.7mph, but there is a bit of a story to that.

I haven't ridden the montour trail at all in about 6 years, so I decided to give it a go. I poked about on google maps for a while and noticed that it intersected noblestown rd out in washington county. I've not ridden noblestown either, but it'll bring me back to the west end bridge... looks good to me.

The first 10 miles to get to the trail flew. Crossing Neville island I was pushing between 20 - 24 mph the whole way. The trail itself is predictably slow. The crushed limestone just sucks energy, and you are riding uphill the whole distance, albeit a very gradual one. There were the usual panoply of kids and families at the start of the trail. The crowds thinned out past Robinson, but the trail conditions also worsened. The climb between Robinson and 79 is long and fairly rough. (rough surface conditions, not rough climb) I was struggling to maintain 12 mph on much of it. Once you get to the plateau at the top the trail conditions get worse yet with a lot of rough gravel, mud, and loose sand. The recumbent is not the ideal platform for such conditions, but I managed. That also killed my averages a bit since I rode a couple miles alternating between 15 and 8 mph due to trail conditions.

Noblestown road on the other hand was a joy. It's a series of nice rollers with almost no traffic until you get close to Carnegie. The climb out of carnegie is long but not bad and there is another climb a few miles later that is about the same. The payoff is the long twisty downhill into the west end which is a lot of fun. You need to dogleg over onto Steuben to get around some construction leading into the west end circle, which I didn't mark on the map, but that's no big deal. Coming over the bridge I started to realize that I hadn't eaten anything on this ride, and that was perhaps unwise. I seem to have a "bonk" ride once a year and this was it.

The clever rider would have stopped at the mini mart at this point and replenished at least enough to get home in good form. I am not clever. Instead I continued on and bonked hard about halfway up brighton. I spent the next several miles watching my average drop from 14.5 mph down to a final count of 13.7. I was struggling to maintain 9mph on flat ground, and was climbing at about 30% of my normal speeds. When I finally got home I almost had to call my wife to help me unlock the garage because my hands were shaking so bad. Stupid, stupid stupid. I do know better. I've done this many times, I know I need to eat about once every 2 hours at a minimum.... oh well.

Overall it was a good ride. I won't do the trail again. It's very pretty, but it's just slow and grueling. Noblestown was a joy though so I'll need to figure out more routes that use it. The bike fealt good overall. I think I need to make a small foam wedge for the seat base since I was sliding down more than I'd like but otherwise it performed well. It just seems to be made for rolling hills and swoopy curves. Once my blood sugar returned to semi-normal levels the only discomfort I have is slightly sore legs.

Monday, April 14, 2008

Seat Building, Finished!

Ok, picking up from where I left off. Once I had the reinforcement on the ribs complete, I needed to put on the final back layer. For this I ordered some knitted fiberglass when I was putting in my order for the epoxy. It's supposed to conform to curves better, and be somewhat stronger, than regular woven glass. It's also a heavier weight fabric than what I had in woven fabrics so I'd only need a single layer for the back. Well it's pretty cool looking stuff anyway.

It does drape reasonably well for such a heavy fabric which is nice. It still took a lot of work though to get it wet out and to lay well into all the compound curves on the back of the seat. This is a process that I think would be better done with a vacuum bag. I probably have 2-4 ounces of extra epoxy in the seat that wouldn't be there if I had a way to compress the material. The finished surface is fairly rough due to the stitching that holds it all together. Using it on a finished mold surface would also be preferable. Having said that I am impressed with this material.

The next step was to apply a couple pieces of reinforcement where the seat mounts would be. In retrospect I think this was totally unnecessary as the finished material is so sturdy. Better to overbuild than have something fail I suppose, particularly since this is really my first experience with the material. (I've prototyped a few ideas in the past with glass, but never followed through with anything)

So the question that plagued me at this point was, how do you accurately mount the seat mount hardware on a surface that curves in all directions. What I came up with was mounting the upper mounts on a threaded rod that is held perpendicular to a piece of square channel AL. The aluminum runs up the centerline of the seat (by eye). And the lower seat mounts are bolted to a block of wood the exact width that the plates are when mounted to the bike, with a cutout for the AL channel. The wood was all cut on the milling machine to ensure accuracy though good skill with a table saw and drill press would have worked fine too. I think the pictures explain it better than I just did. To actually attach the mounts, I used epoxy mixed with milled glass fibers to make a thick paste. I wet sanded (with plain epoxy) the aluminum first to ensure a good bond. I needed a lot of gap filling for the lower mounts since the seat curves so much there. Once that set up I riveted it as well.

I noticed that the knitted glass wasn't totally laminated in two spots on the inside of the rails. What I should have done was drill some small holes and inject epoxy to fill the space. What I actually did was a comedy of errors. I started by grinding off teh knitted glass layer and replacing it with a few layers of regular glass. This was a pain. The following morning I decided it wasn't curing fast enough so I put a hot light on it to speed things up. This worked... a little too well. I proceeded to discolor the epoxy pretty badly. Oh well. It's cured now and it seems strong. I'll just have to live with it. Lessons learned.

Later that day I added the rivets to the now complete seat mounts, and trimmed the seat to width (24cm). I then mounted it on the bike for the first time and sat it to determine where the lower edge should stop. I want to make the total length and shape one of the standard sizes of ventisit pads in the hope of ordering one this summer.

I then trimmed and sanded, and sanded... and sanded ... etc... the seat. A coat of a basic gloss black spray paint later plus some reflective paint on the top half of the seat and I was done! Well, I also had to cut a new seat pad.

Wednesday, April 9, 2008

Wheel Building

So I am swapping out the rear wheel I've been using on the bike for the one from my mtb. It's a pretty decent quality wheel, LX hub laced to a Rhynolite rim. The problem is the rim. It's much too wide for the narrow tires I am running on this bike, so I need to swap the rim. I have a SUN CR-18 rim laying around that I bought on nashbar closeout a year or so ago for a couple dollars. I'm a huge CR-18 fan. They are reasonably light and totally bombproof. I've used rims from other makers and I feel that the CR-18 is equal to any, and costs a fraction. They aren't flashy or aero or anything, but they make a strong long lasting wheel that will take a beating. They also have the same ERD as the Rhynolite so no need to change spokes.

To make the swap, first I removed the casette. This is optional but I'm replacing it too, so it's just easier to do at the beginning. Next I took the new rim and taped it next to the old one with a few wraps of masking tape. I aligned the valve holes and made sure the spoke offsets were going the correct way. I then loosened all the spokes on the side with the new rim and moved them over one by one. I cleaned the spokes and lubed the nipples as I did this. I then went to the other side and started moving spokes over. Start by moving all the spokes that are crossed to the inside of the wheel, then doing the ones to the outside. As I moved the spokes I tightened the nipples so that there was about 2mm of thread showing.

Once all the spokes were moved, I untaped the now free rhynolite rim and set it aside and got to the business of tensioning and truing. I start by tightening all the nipples with a screwdriver so that they just cover the threads. I then went around the wheel tensioning the spokes. I did 2 rounds of 2 turns each for this one. That brought me close to a good tension on this wheel. Before truing I went around and brought all the spokes to even tension by pitch. I then started truing the wheel. I needed to do a fair bit of rounding for this one, the hub was basically off center by about 2 mm. Rounding always takes longer since you need to do up to half the spokes at a time. It also requires more turns to effect a change. I then trued the wheel before checking it for dish. The dish was pretty off (usually is on rear wheels), so I went around tightening the drive side spokes till it was dished properly. I just tightened rather than alternating tightening and loosening to bring the tension higher. I've built quite a few wheels with this rim type so I have a good feel for the tone the spoke should make when it's plucked at the right tension. This one needed to be brought up a little more, so I was favoring tightening.

Once I was happy with the dish, I did a rough truing then stress relieved the spokes by grabbing parallel pairs and squeezing as hard as I could all the way around the wheel a few times. Then I did a finish truing. I put a dial gauge on it, and I'm good within about 0.3mm. Good enough. Lastly I put the new cassette on. It's an 8 sp SRAM 11-28.

All told it took a couple hours working at a relaxed pace and explaining to my 7yo what I was doing.

Tuesday, April 8, 2008

Seat building part III

So the next step in building the seat was to mark out the centerline. I had marked a rough centerline on the mold, but I wasn't too sure about it. So I pulled the set glass form off the mold and put it on the sling seat form and sat in it. I then had my lovely wife mark my spine location directly on the glass. I then put it back on the mold and used the string trick to draw the line.

From that line I marked out the edges of the ribs by marking out lines parallel to the centerline. I think the ribs are something like 6cm from center but I can't remember and I'm not running downstairs to measure right now. One of the things about my design is that the seat stay mounts have to go on the ribs rather than between them. To accommodate this I am going to leave flat sections on the ribs at those spots and the bulk them up with several layers of glass.

To create the ribs I ended up using blocks of pink foam cut and spray-mounted to the back of the seat. I started by ripping strips that were 2" wide by 1.5" tall. I then applied them by mitering them by hand and mounting them 1 x 1.

The next step in shaping them was to use a compass to scribe a line on both sides 1" high. I then used a hacksaw blade to trim the height to this line. I then shaped the foam as best I could using a surform plane and sand paper to a nice round cross section with tapered ends. The spray mount didn't work real well for this BTW, the blocks kept falling off and generally being a pain. If I do this again, I'm going to use epoxy or possibly just spray foam.

I then mixed up some fairing compound using glass bubbles and epoxy and used that to glue down any particularly recalcitrant blocks, fillet the join to the seat, and fill any gaps between blocks. This also didn't work as well as I'd have liked. It was difficult to put it where I wanted it and get it smooth. In the end its OK but not great. Also something I'd do differently next time.

I came back the next day when the filler set up and tried to smooth out the nasty bits using the surform, files, sandpaper, and various dremel attachments. In the end I found the wire wheel on the dremel to give the best control, but it was a pain and ate up the foam if I wasn't really careful. I did mess up a few spots, so I used spray foam to back fill them and then cut and smoothed it out. This worked reasonably well. The spray foam isn't totally consistent throughout, so there are some bubbly bits, but it's good enough to keep the glass in place. Finally a technique that shows some promise!

While the spray foam was setting, prior to trimming it, I cut up some fiberglass. I cut 4.5" wide strips to cover the ribs from a lighter glass (5oz I think). I also cut them on a 45deg bias to allow it to drape better. I cut a total of 6 strips, but I was only able to apply one layer tonight because I ran out of epoxy. Darn. I ordered some this morning expecting this to be a problem, but I won't have it till Thurs or Friday, so I'm stuck till then. Still, things are progressing. I discovered that the bottom of my epoxy container contains a lot of semi-hard gunk. I'm a little worried that the stuff might not be 100% good, so it's probably good that I'm getting new stuff. It seems to have hardened ok so I'm keeping my fingers crossed and we'll see how it goes.

Sunday, April 6, 2008

Seat building part II

Once the plaster set up for about 24 hours on the seat, I scribed a piece of plywood with the shape and made a set of plywood ribs/base. These both stabilize the plaster and give me an easy way to set it on a bench to work on it. To attach them I used two layers of burlap strips soaked in plaster, and just mushed them along the joint.

After that I took the whole assembly downstairs (much to the relief of my sainted wife) and flipped it onto the workbench. I added 3 or 4 (I can't remember, I think 3) more layers of burlap and plaster onto the back of the seat. I am using the mold as a male mold and so I need to account for some amount of seat pad thickness. 6 layers of burlap plus the finishing coats gives me about 1/2" of thickness. I figure this is a reasonable number once you factor in foam compression on the pad. In any case it should get me close enough it won't matter.

After the plaster set up overnight, I went back and started adding layers of drywall compound. I was both smoothing and evening the seat out. There were some odd low-spots and a slight twist to the whole thing which I attempted to correct with successive coats of the compound. The drywall compound is MUCH easier to work with than the plaster for smoothing coats. It stays nice and easy to work for a long time and lets you work at a relaxed pace. Unlike the plaster which is rather frantic. It's actually a very enjoyable process. I was in the throes of a nasty sinus infection while doing most of the drywall compound and I found it quite therapeutic to head down there and sculpt for an hour or so. The other nice thing about the DC over plaster is how easily it sands. I was able to get a very smooth finish very quickly. The drawback to DC is how slowly it sets. The plaster sets up in about an hour. The DC overnight. So it took several days to finish this step.

The last step before applying the glass was to mark a centerline. This was rather harder than I thought it would be. I didn't have the foresight to mark my spine location while the mold was still wet and I was actually sitting in the thing. So instead I just used a piece of string weighted at the ends and laid over the mold and moved it around until it visually looked like it was centered. It's not perfect, but it should be fine.

To apply the glass I started by measuring and cutting out pieces that were about 38"x13". The finished seat length will be somewhere around 35" and the width about 10", so this gives me a bit of wiggle room. I then covered the mold with plastic wrap. In retrospect I should have waxed it and used it directly since the plastic wrap slides around a bunch and is a bit of a pain, but that would have taken a lot longer. For the first layer of glass I weighed it and found it was about 100g. So I mixed 95g of resin with 25g of hardener (100:27 mix for my particular epoxy). I then laid out the glass and carefully applied the resin by pouring it on and brushing with a disposable brush. I applied each successive layer (total of 4) by applying it directly over the previous layer while wet, pressing it in by hand, and brushing more resin on to soak. In all I mixed a total of about 310g of resin and applied about 385g of glass. That's actually a little less resin than I'd ordinarily expect, but it looks good so I'm not concerned about it.

The next step is to build the stiffening ribs on the back of the seat. I'm still figuring out how to do this, so stay tuned.