Wednesday, August 27, 2014

Epiphany


Urban Dictionary seems to summarize it well for me:

"a smart-sounding word for realizing you've been practically retarded for quite some time"

Maybe that's a little harsh, but it's certainly how I felt when I finally realized that I had about 40 hours into building this fuel tank and it's still far from done.

Last night, I brazed studs in the top plate for the filler neck. I had some difficulty getting one of them to completely wet. Funny what a little bit of unidentified trash can do to ya. After that was finally done, I hammered the now warped piece back into shape and prepared to do what I really really hoped was the last couple of steps before welding the top on, namely putting the pickup and return tubes in. I have decided for simplicity to pick up from the bottom rear edge of the tank and return to the top rear edge.

So, I picked up the tank and as I set it on the workbench, I noticed a flash of light inside that was not expected. Ugh. Essentially the entire length of a recently added seam had cracked open. Easy enough to address, and apparently, I hadn't done a good job of it already.

Once the tubes were in, I thought it wise to do a quick leak check. I simply poured some water in the bottom of the tank and watched the seams for leaks as I rolled the water around inside. Found a couple. Fired up the torch and reworked each of them. I let the tank cool to handling temperature and did it again.

Hmmm  How did I miss that leak. Repeat fix/cool/test.

More new leaks. More fixes. More new leaks.


So, I decided to stop the madness.

It seems that either my materials, my tools, my skills or some combination therein is not up to the task. I vote skills. I suspect that my welds are marginal and when working on one fix, metal expansion stresses out new cracks. Maybe the torch is too rich and adding carbon to the weld, making them brittle. Maybe I'm overanalyzing it :)

In any case, there is at least one manufacturer that has an off the shelf tank that would be serviceable, and they take custom requests. I have submitted a request for a quote. Since it's custom anyway, I want to make it as close to ideal as possible, so I have asked for two AN-6 fittings for pickup and return, a marine flange for the fuel level sender, an angled filler neck with vented cap and mounting tabs ready to bolt in place. I might add another mounting tab to mount the fuel pump to; depends on some fitting considerations that I will have to experiment with.

With some luck, I will have a first blush quote soon enough to determine whether the custom tank is even worth considering, which will center around whatever the cost is.

Friday, August 22, 2014

The Tank That Does Not End


The new tank is largely finished, shown here after tacking the panels together and before welding.


I am disappointed with how rusty I was welding. I had to re-weld almost every seam.


You can see how rough and blobby they are here. I spent much of Saturday grinding them clean and retouching all of them. Once I had done a few of them, I seemed to get my groove back.


Had I started with nice welds, I would not have had to grind NEARLY as much. Pay no attention to the pinhole in the middle of the pic. It's gone now.

I have decided to go with an external fuel pump, so I added plumbing for pickup and return. Sharp observers may note that the pickup tube takes off at an angle. That was easier that rebending it or cutting it. Deal with it :)

Also visible in this picture is a baffle added to the tank. It is tacked to the sides of the tank with about 3/8" gap between it and the bottom.


Sadly, it is at this point that I also reach another decision, or more accurately, the legacy of an earlier decision, and it turns out to be a bit painful.

I have opted for external plumbing so that I can remove the in-tank fuel pump and lower the tank in the frame. This will in turn let me lower the body, which needed to be raised to clear the top of tank. By not thinking this option through way back when I drew up the design for the tank, I didn't account for the tank extension interfering with the axle in the lower position. It almost sits on the axle in the lower position.

After much hand wringing, I have decided that the lower tank and subsequent lowering of the body is more important than the extended capacity of the tank. So, I will lob off most of the extension.



The single vertical line is where I will cut. This will still add at least a tiny bit of capacity and will under no foreseeable circumstances interfere with the axle. I had considered moving the bottom of this section up, which would allow only a little bit more capacity and would let me keep the plumbing as it is, but would still have a chance of interfering with the axle.

And to be clear, it's not the axle I presume would be damaged in the collision, but the fuel tank, right there over the exhaust pipe for cylinder 3 and right next to the spark plugs. Lets keep the fuel encapsulated until it's in the cylinders.

In a related design shortcoming, the fuel level sender, as layed out recently, is directly under the same body brace that I need to lower the tank in order to clear. On the plus side, with the new shorter top, I can cut from both the front and the back of the top to make it fit and in the process move the sender forward or backward away from the brace, whichever seems best.

This change will also hurt my plans to make a nice extended fuel filler neck. I will need to simply move the neck off the old tank. I have looked for a new neck. The only ones better than the one I have are weld in parts for aluminum. Maybe the next tank can be aluminum.

This delay has allowed me to come up with a checklist for what is needed to get this trike well on the road and I'd be fibbing if I didn't say it will be tough to make it by Labor Day. They are in no particular order:

1. The fuel tank completed, duh. Finish cutting, welding and lining. Includes new plumbing, cleaning of fuel lines and probably the injectors.

2. Wiring issues resolved; there were a couple of emergency repairs made because wires were so short that they pulled loose from the back of the fuse block and were pretty much unreachable for repair.

3. Wiring diagram; optional but almost a requirement to get all back working.

4. Brake master cylinder reinstalled.

5. Brake slave cylinders replaced (optional, but highly desired; need front wheel cylinders instead of rear)

6. Footpegs painted/powder coated and mounted; will involve cleanup grinding of attachment area on frame tube and possible refitting of mounting flange to tube.

7. Replacement of shifter; most likely, I can just take the shift lever from the new unit and leave the rest of the old one in place.

8. Ignition switch placement. This should go in the fiberglass, but I am afraid it will not be a simple project and may do extensive damage to the paint if not done perfectly. Considering putting it down by the shifter. May end up with temporary alternative, probably using the headlight switch on the control pod.




Thursday, August 14, 2014

More Cutting...

Last night, I rough cut the rest of the fuel tank panels.


I was able to clamp on a straight edge and follow it for all cuts.

I haven't perfected the offset yet, so I do have some parts that might need a little more trimming.


A simple straight edge, clamped at an offset from the cut line, works pretty well. The wire standoff guide is far from ideal, though.

In any case, I have a stack of panels, ready to trim, clean and assemble!

Wednesday, August 13, 2014

The Long Awaited Parting


I have, after MUCH delay, began the construction of the replacement fuel tank for Puff, the dragon trike.

Stacked two 2' squares of 16ga mild steel and plasma cut them simultaneously....

Here is the basic plan:


The layout is intended to use as many of the factory sheared edges as possible to help keep the parts as square as possible. I would love to have used one 5' long 8" wide strip for the top, bottom and ends, but I don't have a bending brake. I have found many inexpensive brakes, but they tend to be limited to 20ga and smaller material. There is a substantial jump in the cost of tools that can bend 16ga. So, instead of about 11 feet of seams to weld, it will be more like 25 feet.

The dotted circles on the top piece are for the filler neck and fuel level sender.

There is a dotted circle shown on the bottom piece is intended for the in-tank fuel pump I am currently using, but I don't like it hanging out of the bottom of a tank that is already pretty low to the ground, so I am considering plumbing this tank for an external fuel pump. On the other hand, I already *have* the in tank pump, it has a pressure regulator attached and it needs no return line plumbing.

Not shown in this drawing are a couple of baffles to be added to the bottom to minimize sloshing. These will simply be panels tacked to the sides and bottom of the tank, arranged and/or cut to allow free flow of fuel, but not totally free movement.


Monday, August 11, 2014

Long Time!

It's been a long time since I've updated here... Been busy with other stuff and even that has been slow to update lately :)

Trike stuff has been slow, but not static. Since that last update, Spongebob has had the most blogable action.

Though it was not immediately apparent, while being stored for a while, Spongebob leaked a lot of oil. Turns out that what really happened is that the fuel tank cap did not vent correctly and the pressure was able to overcome the float valve and flood the engine with fuel, which eventually seeped to the crankcase and overflowed it. This became very apparent when I was rolling the trike out to attend a monthly BTW meeting. It first acted like the battery was nearly dead, which was not unexpected. Unknown at the time, the engine was actually in hydrolock and not able to turn over. Keeping at it, I cleared the cylinders, but when it started up, oil poured everywhere.


Though it would be a few days before I got to it, I drained the crankcase and refilled with fresh oil and started it up. It still leaked oil at too high of a rate to operate, so I presumed the event had damaged the front seal and I again left it for a while.

When our vagabond friend John came to visit, he got interested in effecting the repair, as taking Spongebob for the occasional spin is one of his favorite activities, so I let him tear into it. He replaced the front seal and did a lot of cleanup on the engine, as well as a number of other related work. Even after that, if the cap was properly closed, it would do the same thing. For the time being, we were leaving the cap loose unless the trike was underway somewhere. We also installed an inline fuel shutoff valve. That that seemed to have prevent it from happening at all, though it's not really in a convenient spot to operate all the time.

I opened the cap and reduced the spring tension on the vent mechanism by trimming out a couple of coils. My redneck test of sticking it on my mouth and blowing indicated that I had reduced the required pressure by about half, but in use, it still fails to vent before it blows past the float valve.

I then fashioned a tank vent by putting a fitting in the filler neck and attaching a length of hose to it. There is no pressure build up, so no overcoming the float valve!



In proper followup, I probably need to replace the float valve. Chances are that it may have been damaged by the overpressure situation.

Then comes the day this weekend when we are again wanting to take the trikes to a BTW meeting. I had tried to start the trike the previous weekend when were were attending a birthday party. It would crank and crank, but not start. We were short on time that day, so I didn't do any other troubleshooting. This time, however, I had a little more time and put some more time into it.

I found a badly cracked rubber vacuum cap on one of the carburetor vents. In my experience, that wont usually prevent it from starting, though it can make it run very badly. It took a while to find the package of caps I had, but of course that didn't make any difference. It seemed to be getting spark, but I didn't smell fuel. I started tracing that down and found, surprise, the inline valve was off. Opened it, turned the key on and heard the tone of fuel pump change as it pressurized the system. Hit the start button and he lept immediately to life.

I pulled both trikes over to the house and hosed off the dust; not really a good wash, but roadworthy. We headed in to Fort Worth for the meeting.

As we got close to our destination, the other trike began running rougher and rougher and eventually stopped... in the left turn lane on the access road where we needed to cross over and u-turn to get to the meeting. Try as we may, it would not start. We waited for a clearing in the traffic and I pushed Gabby into a sharp right turn. The road had a bit of a hill on it and she tried bump starting it on the hill and, surprisingly, it started! We were going to try to see if we could make it to the meeting before it died again, but it didn't get far at all. After another try or two at bump starting it again, we gave up and parked it. Gabby climbed aboard and we headed to the meeting, arriving late, but arriving.

After the meeting, a crew of folks went with us to troubleshoot and it was determined that the alternator belt was loose, very loose. Loose enough to prevent it from charging, which apparently let the battery run dead enough to not provide a reliable spark, which seems to be what stranded us. Also why a bump start worked for a bit, with started not pulling the voltage down.

After establishing that there were no shims left to remove in order to tighten the belt, one of the crew produced a spare belt from his parts stock and we installed it. Well, I say we, but nobody made room for me to help, but I let them do it for me. :)

By then, the battery had chemically recovered sufficiently enough to start and the now tight enough belt was actually charging. Long story short, the trike was reliable for the rest of the day and even had noticeably brighter lights. Amazing what you miss if you don't look for it.




Followers