Second practice layup

Posted by lutorm on 1 January 2013, 9:44 pm

After working on insulating the shed and building the hotbox, I got back to the practice layups over the holidays. If you read the first layup post, you know the first try of the 6-ply 10×16″ practice layup was too dry with many white parts. I was hoping that a warm workshop and properly heated epoxy would make the difference.

In laying up the 6 plies, I took care to squeegee the epoxy out over each ply and make several light passes to get the air out. It’s cool, when you pull the squeegee over the freshly wet out fabric you clearly hear a faint sound, almost like fabric ripping, from the tiny air bubbles being forced out. Unlike the previous try, now I added a layer of peel ply over the layup. This helps prevent getting air into the top ply and also makes it easier to squeegee without fear of disturbing the fibers. Unlike the first try, I basically didn’t stipple with the brush at all this time. The result is clearly better than before.

The new layup on the left looks a lot better than the old one on the right. There are many fewer white, dry spots. The weight is 10.9 oz, close to the ideal 11.0oz, but there’s clearly still air in this one too. (The new one looks more opaque because I used peel ply on the back side, so the textured surface makes it less transparent.)

While the first try weighed 9.9oz, this one came out to 10.9, which is very close to the ideal weight of 11.0oz described in the Long-EZ plans. Visually, it looks a lot better than the old one, even if the peel ply changed the surface texture so it’s less transparent than the first one. However, there’s clearly still air trapped within the layup. Holding it up to the light, it’s easier to see.

This is a section of the 6-ply layup in transmission. There are clearly many more air bubbles trapped in the fibers, visible as the zig-zagging darker features.

This is a section of the 6-ply layup in transmission. There are clearly air bubbles trapped in the fibers, visible as the zig-zagging darker features.

The 6-ply layup is a bit large for quick experimentation, so I used some scrap 5×5″ BID pieces to make another, smaller, 3-ply layup. This time I was very meticulous in squeegeeing out the resin until it no longer made popping sounds, and then I started stippling with the brush while heating the layup with the hot air gun. When stippling, the epoxy on top clearly picks up a lot of air, it almost becomes foamy. I don’t know whether this is air driven out of the fibers or if the air gets entrained in the resin when stabbing the brush into the layup. I read somewhere that when stippling one shouldn’t really lift the brush all the way off the surface but rather just lighten the pressure and then push down again. This more careful use seemed to work better, I still ended up with air bubbles in the surface but a couple of quick passes with the hot air gun would pop them. After adding 3 plies like this, I ended up with a layup that looked quite good:

To experiment further, I used 3 scrap 5x5" BID plies and extremely meticulously stippled it while heating with the hot air gun. This layup is close to perfect.

To experiment further, I used 3 scrap 5×5″ BID plies and extremely meticulously stippled it while heating with the hot air gun. This layup has practically no trapped air.

This layup ended up with a resin/total weight ratio of 46%, which would be the equivalent of a total weight of 12.0oz for the large 6-ply layup. This means it’s on the rich side but within the range of acceptability. I could have squeegeed out some resin, but the purpose of this test wasn’t so much to get the right overall ratio as to try to get rid of air within the layup. This was pretty successful. The white features in the picture aren’t from bubbles, it’s the fabric texture on the surface scattering the light, and the large bubble-looking areas are edges on the back side marking areas where air was trapped under the layup, not within it.

Shining the light through the 3-ply 5x5" layup, one can see many tiny air bubbles. (The large bubble-looking things are where air was trapped between the plastic and the layup. It's not air inside the layup.

Shining the light through the 3-ply 5×5″ layup, one can see many tiny air bubbles. (The large bubble-looking things are where air was trapped between the plastic and the layup. It’s not air inside the layup.

Seen in transmission (that’s a pretty cool picture, IMHO) one can still see a few tiny air bubbles, but they are vastly fewer than on the large layup. If it all looked like this, I’d be pretty darn pleased.

So maybe the solution is nothing more complicated than to apply heat and stipple, in addition to numerous squeegee passes? However, stippling is very slow, and the Long-EZ instructions admit as much. I read somewhere that laying up the glass should take no more than 2 minutes per square foot! That is unbelievably fast. At this speed, the entire 6-ply layup should be done in 19 minutes! I probably spent 2.5 times as long. To be fair, some of that was mixing epoxy, and I find that getting resin out in the corners without smearing it all over the table is much, much slower than wetting out the majority of the surface, so it would probably be quite a bit faster if I was just doing a huge surface. However, it seems hard to believe that one could get all the air out and do half a square foot per minute. What kind of layup speeds are people achieving in real work?

Epoxy hot box

Posted by lutorm on 28 December 2012, 5:48 pm

An exercise that almost everyone that starts working with fiberglass encounters right away is that of building a heated box for epoxy storage. The epoxy should be warm when used, somewhere around 30C or 90F, otherwise its viscosity is higher and it’s more difficult to wet out the fabric. If you don’t store it at that temperature, you’re faced with heating up the gallon-sized container of resin from its storage temperature before you can start working, so most people build a small heated box where the epoxy can live. (I think a second reason is that the hardener can crystallize if it sits at too low temperatures. You can recondition it by heating it up, but you may not notice that it’s started to gunk up your epoxy pump which will screw with your resin to hardener ratio.)

I decided to build a simple 2’x2’x1′ box out of plywood and insulate it with the same kind of styrofoam sheets I’ve been using to insulate the work shed. The inside would have two compartments divided by a shelf, a small one on the bottom where the heating element was and the main epoxy storage compartment on top. I just put the whole thing together pretty quickly without much thought to appearance, but I did decide to laminate the 6mm plywood shelf with fiberglass, not so much because I thought it wasn’t strong enough but because I didn’t want it to suck up the inevitable epoxy spills. It would also give me a chance to try laying up epoxy on wood.

Laying up the hotbox shelf. The holes are so air can circulate through the two compartments.

The most obvious thing from this exercise is that wood sucks up epoxy like crazy. I’d wet the wood, then put the single ply of BID on there, and then after a minute it would look totally dry. I’d add more resin, then the process would repeat. It would have been interesting to weigh it to see how much resin it ended up infused with.

I also added a ply of BID to the exterior bottom of the box and across the edges since the plywood was only attached in the corners. This gave the box some extra stability and me an opportunity to practice making corners, filling gaps with micro, and peel plying the edges to avoid sharp glass needles poking up out of the surface.

The exterior of the hotbox. I didn’t put much effort into making it pretty, obviously.

So far so good. Time to do the heater. It seems a lot of people just buy a thermostat and connect it to a light bulb inside the box. This would work, but seemed way too simple… Those of you who have looked at my old “projects page” on my home page may have seen the steam boiler monitoring system I built when we lived in Massachusetts. When we moved, I tore out the temperature probes and the box, and this seemed like a prime opportunity to reuse that stuff. It would also make it possible to add temperature control of the shed itself at some point.

The old system used a pair of “Series 1” XBee radios to send the temperature data to our server. These didn’t have enough juice to cover the distance between the shed and where we have the server currently, and then there’s the additional complication of the shed being made out of sheet metal, so an external antenna would be needed. Since none of my existing radios could use an external antenna and I didn’t even know if they would work, I got 2 new Xbee “Series 2” radios. These are quite cool in that they will generate a mesh network and route packets between them, so even if the destination is not reachable from the source node, the data will get there as long as there’s a chain of nodes that can talk to each other. Thus, if I had problems getting reception, I could add another node somewhere in between and it would act as a router.

It turned out they managed to talk to each other fine, even from within the shed as long as a door was open, so the external antenna I got must have quite a bit more gain than the little “chip” antenna the old one had. Still, it needs to work with the doors closed, so I mounted the antenna outside the shed.

Since it was hopeless to get the XBee to work from within the sheet metal shed, I got a 2.4GHz antenna with a 10ft cable and mounted it on a piece of scrap plywood above the shed. One of the outside temperature probes also lives on that mast.

For controlling the temperature in the hotbox, I got two 2 Ohm, 25W power resistors at All Electronics (it’s nice to have found a good DIY electronics place in LA, even if it’s quite a drive). This is controlled through the Arduino PWM with a power MOSFET mounted in the Arduino box. With the resistors in series that makes 3A or 36W at full power. I mounted two of the old DS18B20 temperature probes from the boiler inside the box for temperature monitoring, one more on the power resistors themselves (to detect if the fan fails) and one to measure the ambient temperature outside the box. Two more temperature probes (the old living room ones with long wires) were pulled outside the shed to measure outside temperature, one where the radio antenna is and the other on the north side of the shed that never sees sun.

This is the enclosure that holds the Arduino, XBee radio, and the connectors for the hotbox heater and temperature probes.

The heating elements for the hotbox are 2 2ohm, 25W power resistors connected in series. An old computer fan is used to circulate the air over them.

The heating elements for the hotbox are two 2 Ohm, 25W power resistors connected in series. An old computer fan is used to circulate the air over them.

The heating element is controlled with a PID controller on the Arduino, with an added open-loop control based on the temperature outside the hot box. Since the temperature inside lags temperature changes outside, the controller performance is enhanced substantially by the open-loop part. (This was used in the wine refrigerator, too.)

The software on the Arduino and on the server didn’t need many changes to work with this new system. I did have to do some coding to be able to upload new firmware to the Arduino over the radio link because the Series 2 XBees doesn’t quite work exactly like the old ones when it comes to controlling digital output pins on the remote end, which is needed to reboot the ATMega and start the boot loader. Basically, instead of just sending data straight over the serial line to the radio, you have to assemble them into packets depending on what you want it to do (send the data out over a remote radio’s serial line, change a setting, change a setting on a remote radio, etc). It’s more complicated but gives you greater control of what you can do.) Anyway, I worked on that over Christmas, aided by the book “Building Wireless Sensor Networks”, and that now seems to work reliably.

The inside of the hotbox. The walls are insulated with styrofoam insulation, the epoxy sits on the shelf, and the heating element is on the bottom. The Arduino controller box sits on top.

The system seems to work well. After putting it out in the shed and adding the epoxy containers, it maxed out the heater to heat them up, which took about 5h. By then it was getting colder and the power was going up until it maxed out when the temperature in the shed hit about 10C. (It got down to 4C outside the shed — I thought Southern California was supposed to be warm!) Then when the sun rose, the shed rapidly warmed up to 20C and the heater dropped down to 40%. It doesn’t quite have the power to keep the box at 30C, but as you can see above there’s no insulation around the hinges so there’s some room for improvement. Or it might be good enough, I doubt I’ll be doing much work when it’s 4C outside anyway.

The temperature data are shown on plots generated by our server. The plot below shows the temperature of the hotbox and the shed over the last 24h, updated in realtime:

The hotbox temperature over the last 24 hours.

Now maybe I can get around to making the second try of the test layup…

NC30 front brake replacement

Posted by lutorm on 13 December 2012, 12:36 am

The latest NC30 update is pretty pedestrian: it got new front brake disks and pads. The old disks were below the service limit and the play in the “floats” was enough that the wheel could move noticeably with the front brake held in. The pads were also hopelessly oily from the leaked fork oil, so even though I cleaned them out, the disk would get oily immediately. Time for a replacement.

I got 2 new brake disks from EBC and new pads. For such few parts, it was one of the most expensive maintenance items so far, the pair of disks were over $400… I hope I don’t have to do that often. Happily, there were no snafus on this one — the new disks bolted right on, the old pads came out, no sign of bad seals so I wiped out what gunk was around the pistons and put the new pads on. While I was at it, I flushed out the old brake fluid, too. The most exciting part was lifting the front wheel up by hoisting the bike with a set of straps around the triple clamp and a ceiling beam!

The new brake disk, looks spiffy with the gold colored center.

The new brake disk looks pretty spiffy with the gold colored center.

After taking it for a test ride, I’m very happy with how they feel. On my old Ninja 250, putting a new disk on resulted in basically no braking power at all initially until the pads had worn in. Here, I just did a couple low speed stops going down the street and they were feeling perfectly functional. (Maybe that’s what you get with 8 pistons instead of 2…)

Now it should be perfectly road worthy. I was looking forward to taking it up to SpaceX for the first time, I’ve been keeping a running commentary on bike restoration with some other bikers there so it would be nice to finally get to show them the result. However, there’s one little remaining issue: I need a new headlight bulb. The headlight is some Tyga special, and the bulb is weird 35/35W I’ve never seen before. It’s apparently common in Asia, but doesn’t seem possible to get in the US. I just ordered one from Tyga in Thailand (they even offered free shipping!). The headlight is pretty much a piece of shit, it’s poorly attached and has no aiming capability, but there’s not much I can do about that without replacing the top fairing. I’m not quite ready for that yet, there’s plenty of other things on the plate…

NC30 first ride

Posted by lutorm on 25 November 2012, 7:57 pm

With the carbs and valves finally done, it was time for a little test ride. (After having the bike for over 3 months, I’ve so far basically only ridden it around the block twice to test the forks after rebuilding them.) After putting the fairings back on, I shot some pictures and went for a ride.

I shot a video of it starting up. It’s obviously not the first start, but I still thought the situation warranted saving it for posterity.

After riding at maybe 15km around the local streets and a half-mile stint on the freeway between Artesia and Crenshaw, some observations:

The engine runs really nice, it pulls cleanly with increasing power from pretty far down all the way up and idles steadily hot or cold. It sounds a lot better at 13-14k rpm than the NC23 does. At least something is right!
It’s loud, even with the muffler insert. I’m ambivalent about this, on the one hand it sounds kind of aggressive, on the other hand I really don’t like people riding around making loads of noise. And forget about riding it without earplugs.
The clutch is really heavy, even with a new and freshly lubed cable. My RSI-scarred hands were not happy about it after a while.
After heating up, the clutch also drags a bit even when fully in which makes it hard to find neutral. This may be a NC30 “feature”, it seemed quite a few people on the 400greybike forum have the same issue and recommended blipping the throttle up to maybe 3000 rpm, and this definitely helped.
The gas is also a bit heavy. I guess I just need to build up some hand strength.
The seating position feels quite different from the NC23. It feels “shorter” somehow, like you reach less forward and more down with your arms. It’s also noticeably lower and narrower.
It handles nicely. This is what everyone says about the NC30, and it seems true. In the few turns I went through, it felt willing to turn in and stable through the turn.

That was fun, but before doing any major riding, it needs new front brakes. The disks are about 0.4mm below the service limit and there’s major play in the floating disks.

NC30 Carb overhaul part 2

Posted by lutorm on 25 November 2012, 1:22 am

The first post about the carb overhaul ended with “now I just have to synchronize them”… Well, nothing ever seems easy on this bike.

For those of you who don’t know, synchronization is the procedure by which you make sure all cylinders are idling at the same throttle. If you don’t do this, chances are that a few cylinders will want to go faster and will have to pull the others along. This results in a bad idle and the cylinders that are doing all the work will run hotter than the others. The procedure is to hook up a set of vacuum meters to the intakes on the cylinders. Then, with the engine idling, you adjust the screws in the throttle linkages between the carburators until the manifold pressures are the same.

This is simple, in principle. The problem starts when you try to do this in practice. There are 3 adjustments screws that each adjust the #1, #3, and #4 cylinders against the #2 one. These screws are all between the carbs, facing down. To get to them, you either have to jam your hand in between the engine and the carbs, which is not going to fly since this is done with the engine at operating temperature and you’ll burn your hand badly, or you have to come up with some creative tools to allow you to reach the screws. (I’d show a picture but it’s impossible to capture.)

Once I’d worked out some way to reach the screws, I started the adjustment. I had problems with the #1 screw. I wasn’t sure I was turning it the right way, because it seemed to not have the result I expected. Eventually, the screw became really loose and when I tried to screw it back in, it popped out. Sigh… After trying to get it back in, I bit the bullet and took the carb assembly off. It turns out that this screw was not original. The standard screws are M4x0.5mm fine pitch, but someone had jammed a standard M4 screw in there, and thereby messed up the threads. This is a 2mm thick metal piece with not a lot of room.

I pondered my options. I first thought the thread was good enough to hold the standard screw in there, so I ordered a new original screw. Alas, when I tried to screw it in, it popped out under the spring pressure. No go. Next I wondered if there was some way to repair the thread. The material is too thin to use a thread repair insert like a Time-Sert. I thought about trying to repair it with JB Weld and re-thread the standard screw, but threading in JB Weld, in such a thin material, was also unlikely to last. The last thing I wanted was for the thing to pop out after a few days, necessitating yet another carb removal.

In the end, I opted to drill up the hole and tap a new M5 thread. This seems to work, knock on wood. I could only find standard pitch M5 taps and bolts, so it’s a rougher adjustment than it should be, but that’s life.

After putting the carbs back on, I tried to yet again synchronize. I was getting the hang of how to get to the screws, so it was a lot faster than before. However, the results weren’t making sense to me. The Haynes shop manual says that the three screws adjust #4 to #2, #1 to #3, and the front (#2/#4) to back (#1/#3), but it seemed to me that when I turned the screws, something different was happening.
I posted a question on the very useful 400greybike forum, and someone responded “that’s your problem: it doesn’t adjust front to back”! Apparently the Haynes manual is incorrect on the procedure…

This movie is toward the end. I did a few more adjustments, but this is pretty good:

Now we’ll see how it runs on the road! It would be nice to finally get to take this thing for a ride, but we’re not quite there yet — it needs new front brakes first.

First fiberglass test layup!

Posted by lutorm on 20 November 2012, 1:26 am

After the SportAir workshop, I put in an order for a startup kit of composite materials from Aircraft Spruce. Since we got a discount from the workshop, I thought it would be a good way of getting some momentum on the airplane building. I didn’t order nearly enough for an airplane, but it’s sufficient for starting to practice with the materials.

The first practice layup that the Long-EZ plans wants you to do is a flat rectangle made up of 6 layers of bidirectional (“BID”) fiberglass. This gives you practice in wetting out the glass, making sure you get all the air out, and to not oversaturate the layup with resin which would make it unnecessarily heavy.

By now our shed in the back is pretty cool (it’s just made of sheet metal without insulation) so the temperature was around 18C, a bit lower than ideal. If the temperature is too low, the epoxy resin becomes thick and it’s harder to get it to penetrate into the glass. This was definitely noticeable as I spread it out on the table against the probably even cooler bench top, I had to work quite a bit to get the epoxy into the glass. This is done with a squeegee (carefully so as to not disturb the glass fibers). As there were numerous air bubbles visible between the glass layers, I used the “sandwhich” method, putting a layer of plastic over the top. This makes it possible to squeegee harder without disturbing the weave, and with generous use of the heat gun and dragging the air bubbles out to the edge, I got a pretty nice layup.

This is the 10×16″ practice layup, consisting of 6 layers of BID. It’s supposed to weigh 10.5-12.5 ounces, mine came in at 9.9 so it’s too dry. This is also clear from the numerous white areas where the fibers are insufficiently wet out with resin.

The picture above shows the cured and trimmed piece of glass. The white parts are where the fibers are insufficiently wetted out with resin. I noticed them already as I was finishing the layup but decided to let it cure and see how it looked. Either the temperature was too cold or I squeegeed out too much resin from the plastic sandwhich which allowed air to get drawn into the fibers when I removed the plastic. Most probably, both were issues. According to the instructions, with an ideal amount of resin, the 10×16″ piece should weigh 11 ounces (yeah, I know. That’s 311g in civilized units) with a weight between 10.5 and 12.5 being acceptable. Mine weighed in at 9.9oz, another sign that it’s too dry. (For reference, the glass cloth alone weighs about 6.5oz.)

A closeup of a part of the glass. The white is air within the fibers, not good.

Apart from the dry spots, the piece came out pretty well. It’s surprisingly strong. The piece is about a mm thick and with considerable effort I can bend it 90 degrees and it will then spring right back.

I’m going to try making the piece again, but before that I’m going to need better temperature control in the shed. I tried putting a space heater in the shed to get the temperature up a bit, but the bare metal just sucks the heat out and the power bill would cost a fortune. I’ll try putting up some insulating styrofoam sheets on the inside, even a thin layer should be a dramatic improvement over the current state of affairs.

I’ll probably also need to make a heated epoxy storage box so it can be stored at 30C or so. People commonly just make an insulated box and mount a light bulb with some form of temperature control. I still have plenty of DS18B20 temperature probes and a spare Arduino so that should not be a big deal. Then I think I need to build a large table, too. The work bench is fine for these practice pieces but for the real deal I need more surface area.

NC30 Valve Adjustment

Posted by lutorm on 10 November 2012, 12:46 am

Along with the carb cleaning, checking the valve clearances on the NC30 going to get done while I had stuff apart. While access is tricky (you have to remove the radiators and pretty much the airbox as well), it was actually easier than on the NC23 because you don’t have to remove the camshafts to change shims. Instead, you just slide the rocker arm off the shim and pick it out, which can be done (after a small amount of swearing to bend your fingers into the right shape) in a few seconds.

One thing that tripped me up was that you are supposed to measure the clearance between the shim and the rocker arm, not between the rocker arm and the cam. It seems physically impossible to me that you’d get different results, since you just push the rocker arm either up or down, but fact is that you do. Measuring between the cam and the rocker gives about .07mm smaller clearance than the correct measurement. Ask me how I know…

The funny thing was that once I measured the correct clearance, they were all too large. Valve clearances go down with time as the valve seat wears and the valve sinks deeper into it, so large clearances can really only happen if there’s crap between the valve and the seat. In that case, the valve won’t seat properly and your compression will suck, but my compression was good. And this was pretty much even across all valves… so I suspect whoever did the previous valve adjustment also measured between cam and rocker and set them all to too large values… It’s like the bike mechanic equivalent of the Keystone Cops has gone over this thing…

The most annoying thing about the doing the valve clearances is that you need to order the correct shims and then wait for them to arrive. And everyone pretty much only covers the .05mm increments, except Honda who has them in .025mm. So if you end up needing the uneven sizes, you can’t even get them from a local shop. The bike pretty much sat around 2 weeks waiting for shims.

But anyway, they all got adjusted. For good measure, I got new valve cover gaskets since there were signs of previous oil leakage in that area. I also looked over the cams, which looked fine with one exception:

Everything looked good except the inner exhaust cam on cylinder 3, which was scored as seen on this picture. The thin scratch marks were made by me with a screwdriver because I wanted to feel if it was actually a noticeable pit, which there is.

The picture above shows one of the exhaust cams which has a longitudinal black mark on the lifting side where the pressure is the highest. I scratched it with a screwdriver and could actually feel a pit in the surface, so it’s not just cosmetic. Hopefully this is not the beginning of something progressive.

With carbs back on, valves fixed, and a new thermostat fitted (the old one was stuck wide open), I attached a funnel to the fuel line, poured enough in to fill the carbs, and cranked it over. Not a sign of life… Man, what could it be?

Right, kill switch is off… Put it back on “run” and it started right up! Now the final thing that needs to get done before trying it out is synchronizing the carbs. Getting access to the throttle adjustment screws, which are between the carbs on the under side, is not going to be easy. Whoever decided that was a good place should have his head examined…

NC30 Carb overhaul

Posted by lutorm on 6 November 2012, 12:07 am

Since the NC30 wasn’t idling on all cylinders and generally was misfiring at low rpm, I figured it wouldn’t hurt to take the carburetors apart and make sure everything was cleaned and adjusted right. This turned out to be a good decision… While the insides were pretty clean, things were pretty garbled. There’s a good guide to NC30 jetting on 400greybike.com, and it points out that several parts are different between the front and rear carbs so it’s a good thing to check that they haven’t been swapped. Mine were…

The first thing I noticed were that the main jets were #110, which isn’t right. Only the first gen ’89 NC30 had #110 main jets, the later models should have #115 front, #118 rear. The numbers on the carburetors also matched that they were from a ’89 model. Since my NC30 is a later ‘N’ model, I can only assume that someone swapped the carbs at some point and didn’t know the later models had different jetting. This didn’t start promising. The next thing the guide says to check is the needle and needle jet, as they are also different. Sure enough, my front carbs had one front and one rear setup.

At this point I’d decided that I was going to disassemble the carburetors, something they generally don’t recommend unless necessary. Well, it was necessary: there are short pieces of tubing between them for the air pipes, and those were just totally coming apart. They weren’t the original pieces, either, but some generic hose pieces held on with a zip tie. Since you really don’t want hard-to-find vacuum leaks in this area which takes a lot of disassembly to get to, I decided to just start over. A lot of the O-rings were quite stiff so I even decided to buy a new gasket set. These aren’t easy to find and cost their weight in gold. (Probably a lot more than gold, actually.) I found a set on ebay that shipped from Japan.

So far so good. Then I discovered the following while doing some general cleaning:

This hole was hidden under a bunch of gunk. How do you do that??

There’s a metal cover at the end of the throttle axis, and after getting a bunch of gunk out of that area, I discovered an actual hole straight through the metal! How do you manage to do that? This actually goes through, you can see movement through the hole when you turn the throttle, so it’s going to have to be sealed up or there’ll be another vacuum leak. The gunk in that area was probably epoxy or something that some previous owner had “fixed” it with. I guess I’ll have to cover the hole with JB Weld or something. It’s supposed to be fuel resistant so hopefully it’ll last.

Finally, the air hoses were cracking, and rather than try to find new ones for something like $20 each I decided to get some silicone vacuum hose from Hosetechniques.com. I used their vacuum hoses in the Passat and after 7 years they look like new, plus I realized they are practically down the street from here!

So finally, this is the reassembled stack, ready to go back in the bike. The blue hoses are the new vacuum hoses, and I also used short sections of the same hose to replace the ones that were coming apart. They’re visible between the carbs on the left in the picture.

Here’s the assembly put back together and ready to go back in the bike.

While I had everything apart, I also replaced the throttle, choke, and clutch cables which were in bad shape. The new ones definitely feel better. Getting the carbs remounted was a bit of a challenge. Each carb in connected to the intake with a short rubber tube, and since the engine is a “V”, they aren’t all parallel so getting them in takes a bit of flexing of the intake rubbers. Luckily I also had gotten new ones of these as the old ones looked cracking and were also more like hard plastic than rubber. The new ones were a lot more flexible so I’m sure it would have been a lot harder without new ones.

The carb assembly back on the bike. Getting them to seat properly in the intake rubbers was a bit of a challenge, but at least these were new and pliable.

Everything is now mounted back on the bike and awaiting test. In parallel, I’m adjusting the valve clearances and I’m waiting for some shims to be able to finish that. The front cylinders are done so it’s safe to remount everything. (Accessing the front camshafts is a lot harder than the rear. You can see the rear valve cover plainly exposed in the image below.) Once that’s done, a carb synchronization is in order and then I hope it’ll run like new!

NC30 Fork rebuild

Posted by lutorm on 31 October 2012, 9:14 pm

I’ve been collecting parts to start working on the NC30. The first step was to fix the leaking fork seals. This turned into a bit of a longer project after taking the forks apart and noticing that the slider bushings were in pretty bad shape, too, so I figured I should just replace them while I had the whole thing apart.

This is what the old fork seals looked like, split and rusty. I don’t know what could make them look like that, except maybe if they were driven in with a screwdriver or something that damaged them. Hopefully it’ll be a while until the new ones look like that…

The fork bushings were in pretty bad shape, too. This was admittedly the worst one, but there were clear signs of abrasion so I figured they should all be replaced as well.

While I had the forks off, I also disassembled the steering head bearings to clean them out. They were in decent shape, with this glaring exception:

One of the steering head bearings. Note the huge flat spot on the ball in the center! This must have been preexisting, because there’s no way you could do that to the ball without damaging the races, which had no signs of abuse.

I don’t know how you make a ball look like that, but I suspect it must be a manufacturing defect since all the other balls and the races look fine. The top dust seal was torn so I replaced that too. (It wouldn’t have hurt to replace the bottom one but you can’t get that one out unless you remove the lower race which I wasn’t about to do.)

Finally, I also replaced the dust seals on the front wheel bearings which were pretty crusty (and had been painted over which I’m sure doesn’t improve their flexibility)…

After putting it all together and taking it for a short test drive, everything seems to be fine. No noticeable play in the steering head, no noticeable stiction in the forks. I still have to tune in the fork preload and damping, but it’s probably easier to do that when the bike is fully running…

EAA SportAir Composite Construction Workshop

Posted by lutorm on 30 October 2012, 11:44 pm

The Experimental Aircraft Association has been running these workshops teaching different aircraft building skills around the country for a long time, and when I saw that the Composite workshop was going to be offered in Riverside, I jumped on the opportunity to get some momentum on my airplane building project. (It was apparently a good idea, too, because our instructor, John Brecher, said this was the second to last one. From now on they’re only doing them in Oshkosh…)

We worked on two little pieces, a flat piece of foam we glassed, cut in half, and then joined together with a fillet joint and a small section of an airfoil that we hot-wired in polystyrene foam, added spar tapes top and bottom, and then skinned with fiberglass. This took two full days, mostly because of the need for the epoxy to cure for 24h between each layup. I’ll let the pictures speak for themselves below, but it was a nice introduction to the techniques used and it doesn’t seem that hard. Of course, like I said at the end of the workshop, being able to manufacture some pieces is one thing, trusting them with your life is another…

: This is the resin we used, West system 105 + hardener.

: This unidirectional fiberglass tape is used when you need all strength in one direction, like the spar tape next.

: This spar tape is made up from multiple layers of the UNI tape on the previous picture. It’s maybe 5mm thick and completely rigid in the fiber direction, while noticeably flexible in torsion.

: This is the foam we used. The gray sheet is polyurethane, the blue polystyrene.

: The first step is to apply micro slurry to the foam surface. This fills the open cells on the surface of the foam and ensures a good bond between the glass and the foam.

: The micro slurry is scraped into the foam cells with a squeegee, ensuring that only a thin layer is left on the surface.

: Then the fiberglass cloth is added, and the epoxy resin worked into the fibers so that there’s no air bubbles. (I haven’t done such a good job here.)

: The final step is to add “peel ply”, a layer of Dacron fabric which the epoxy doesn’t stick to. After the expoxy has cured, this can be peeled off (hence the name), exposing a fresh, textured surface for bonding additional layers of fiberglass to. This saves you from having to sand the surface to get a good bond.

: On the second day, we removed the peel ply, exposing the fiberglass.

: The object was now to make a T-joint. The first step is to just fix the pieces with some hot glue to hold them in place while the glass is being applied.

: This is the final joint after applying peel ply. It was a bit messy and I didn’t want to get epoxy on the camera so there are no pictures of the intermediate steps. Basically you first fill the corner with thick micro to make a radius that the cloth can follow. Then three layers of glass of increasing size were added over the corner, followed by the peel ply.

: To cut the airfoil, we started by mounting the templates, one on each side of the foam core, making sure they are parallel.

: Then you cut the foam using a steel wire stretched between two posts, heated to red hot using a variable transformer. By sliding this along the template on both sides, you cut the exact shape into the foam. If all goes well…

: Here, the bottom has been cut and they are working on the top.

: This is what I ended up with. It’s pretty good, but you can clearly see some artifacts where we weren’t quite at the same point on both sides. Staying synchronized so you don’t cut diagonally is the big challenge.

: The top surface came out better. The “tab” on the left is the trailing edge. It will later be cut away, but not until the bottom has been glassed.

: The first step is to lay up the thick spar caps on the top and bottom using the unidirectional tape.

: Then we cut these pieces of BID that will become the surface skin.

: Here I’ve done the bottom spar tape and skin. The glass looks a bit dry.

: This is the trailing edge. When we cut the foam “tab” away, the top skin will join this part, making a sharp end.

: On the leading edge we put peel ply where the cloth ends. This makes the fibers lie nicely flat against the surface so the other side will overlap cleanly.

: After curing, the excess glass can be trimmed with a handy multi-tool.

: The leading edge after removing the peel ply. A small ridge of resin is visible that will have to be sanded down.

: Here the top side of the airfoil has been laid up with spar tape and skin. Note that the foam “tab” is now gone and the trailing edge tapers to a thin edge.

: The final technique that was demonstrated (we didn’t make one each) was vacuum bagging where you cover the object in plastic and pull a vacuum that “squeezes” excess resin out. Here we’re starting with a flat foam piece with a little shaped object in the middle.

: Then you add epoxy normally.

: Though you don’t have to worry so much about getting too much epoxy in the fibers, since it will get squeezed out.

: Add peel ply on top as usual.

: This “padding” is a thick material that will suck up excess resin and enable the vacuum to get everywhere across the piece.

: Then the entire layup is covered with plastic.

: The plastic obviously needs to be totally sealed. The yellow is a very sticky double-sided tape that’s used to attach the plastic. The little pipe poking out is where the vacuum pump is attached.

: When the vacuum pump is turned on, it sucks the plastic tight against the surface, making sure there are no air pockets in the epoxy or between the glass and the foam.

: After sealing the piece with some extra duct tape visible in the background, the vacuum was up to 24 in Hg, which is sufficient.

: You can clearly see how well the cloth is conforming to the shape of the foam piece. Unfortunately I don’t have any pictures of the piece after curing, because it didn’t really have time to cure fully before the workshop ended.

Patrik's projects

A random smattering of ideas turned into reality