In our last episode, things were imploding and our fearless leader was contemplating how he could save the world, er, mold.
Recall, things started imploding as leaks were present between the outer skin of the plug and the inside space.
To try and fix things, I cut out the bottom of the
Masonite base to access the hollow interior of the plug. Applied some sealant around the inside corners to help get a good seal. Then I poured in some "dump foam". This is the stuff that comes as a two part liquid. Mix equal amounts, and pour. Within ten or fifteen minutes, the liquid expands many times its volume and cures to form a lightweight closed cell foam. The dump foam gives the mold strength, and helps seal any area
that may be the source of leaks. So, once filled, I applied the vacuum. And voila! I have -28
kPa pressure (pretty good for this vacuum pump). There was still a very tiny leak somewhere that I could not seem to locate, but I rolled the dice and gave it a go.
Things seemed to go pretty well at first. However, after a while "raceways" were developing along the edges. "Raceways" is the name given to channels that develop where bubbles all of a sudden "zip" toward the vacuum port under pressure. This indicates where resin is freely flowing. The effect of this is as the glass wets out, the resin will travel from the feed tube, to the vacuum tube, along these raceways (a path of least resistance), before other areas of the part are fully
wetted out. The only thing to do is to continue drawing more resin through and hope those areas finally wet out.
The next morning, I pulled the bag, flow medium and peel ply. Sure enough, thought not totally dry, parts of the inner layer of glass was starved of resin. For further inspection I started to remove the part from the mold. While the laminate seemed to separate from the surface pretty easy, it was hard to get a handle on things to pull apart. So, being as this is a one-off part, I
began to cut the mold into pieces from the inside with a
sawzall.
Success? Well, Sort of...Well, so the infusion did not come out as planned, and the piece as it is, cannot be used as a tank. And I must say, this is one of the more "complex" pieces I have infused so far. But it is not a total loss.
The important part here is getting the odd shape that fits the space where it is intended to go. That is the whole reason for making a custom tank here in the first place! And, since it is just a poo tank, it simply has to be solid and well sealed. Not perfect.
Two Salvage Options
If I really want to pursue and infused tank, I could use this first piece as a FEMALE mold for another attempt at the tank. I would need to do some hand layups of a couple layers at least inside. Then I could smooth it out and try infusion again. And throw the mold away.
Or, I could do the hand layups. Make sure it is sealed well, put a top on it. And have a tank! Not wanting to dwell on the infusion aspect too much, I might opt for the latter and simply finish it with hand layups as a tank.
Lessons from this Project...The
Colbond flow medium I used (blue random stranded material layered outside the peel-ply) should not be used around the edges of the part, as when pressure is applied, it compresses the medium inside the bag and creates a small channel that runs along the edges. This is what created the "raceway" which helped the resin reach the
vacuum port before other areas of the part were saturated. I should have cut the flow medium only to cover the face of the sides up to about an inch from the edges.
Next time, I will used multiple
vacuum ports, with valves, which will enable a degree of control should one part saturate faster than another. Ports could be
controlled to help "encourage" the resin to flow to other areas of the part.
Larger
vacuum lines. I used 1/4" ID hose this time. Next time I will use 3/8" ID.
Continuous strand mat. I was thinking about using this over the final layer of glass, but decided against it. In addition to adding strength, it would have helped resin flow. I will try that next time.
Got some more foam cut and glued this weekend. Actually, all but one piece of foam has been shaped. But I ran out of glue. More coming today. Also, the bottom-most stainless "webbing" piece, got a "sacrificial skid plate" attached to the under side. Made of out of 3/4" Coosa and 1/4" G-10 plate (bottom most) the idea being, in case of some sort chafe from grounding, the Coosa, G-10 and the glass laminate that surrounds it, will take the hit first before the stainless frame.
Yes, fairweather comes to the PNW. At least for a couple days. Rented a 3600 PSI pressure washer and took it to the Westsail, and other things around the yard. This washer was more powerful than the one we used last year. Unfortunately, it was not powerful enough to remove the gelcoat from the deck, which I was kind of hoping (that is a joke by the way, I am still NOT looking forward to grinding off the waffle pattern nonskid from the deck.)
Also, the tape I had been using to seal the hull/deck joint to keep the water out had been failing. The UV ultimately got to it. I was surprise it lasted as long as it did. Unfortunately it left a lot of sticky adhesive residue which the pressure washer could not remove. Not a big deal at this time. But I do need to devise an alternate method for keeping the rain out before it comes. 
Got the final rudder welding done over the weekend. Many thanks for Bob and Jan over at Fidalgo Machine, who, not only did the final welding, but also machined many of the bits. Special thanks to Bob for putting up with my questions about all the machining he does. I would love to have a shop like that to play in. If I only had the time.
As mentioned in previous posts, we are planning to install transmission steering. One of the other advantages of transmission steering, (over pull-pull cable or hydraulic steering) is the mechanical advantage that can be achieved using wide angle geometry. Basically, the principle lies in the difference in lever arm lengths between the reduction gear box (right in the picture) and the rudder post tiller arm (left). The reduction gear tiller arm can swing through an arc of about 120 (60 degrees port and starboard) degrees. The typical 'useful' full swing of a rudder is about 72 (36 degrees port and starboard) degrees. By aligning the reduction gear, with the tiller, such that are linked in the "full swing" position, something mechanically interesting happens. 
Recall, things started imploding as leaks were present between the outer skin of the plug and the inside space.
Things seemed to go pretty well at first. However, after a while "raceways" were developing along the edges. "Raceways" is the name given to channels that develop where bubbles all of a sudden "zip" toward the vacuum port under pressure. This indicates where resin is freely flowing. The effect of this is as the glass wets out, the resin will travel from the feed tube, to the vacuum tube, along these raceways (a path of least resistance), before other areas of the part are fully wetted out. The only thing to do is to continue drawing more resin through and hope those areas finally wet out.
The important part here is getting the odd shape that fits the space where it is intended to go. That is the whole reason for making a custom tank here in the first place! And, since it is just a poo tank, it simply has to be solid and well sealed. Not perfect.
