Friday, March 23, 2018

Fixing a fuel tank baffle leak

As I mentioned in the previous post I had a baffle leak on the right fuel tank.  This is the process I went through to repair that leak (two actually).

First step was to cut two 5" holes in the middle of the leaking bays of the back baffle.  Here I set the fuel tank on a blanket on the floor and then used the 5" hole saw you see sitting on the bench (top down view) to cut the holes.  Pretty simple process but the cuts definitely needed some cleaning up.
Once I had the tank open I was able to look inside with a mirror and see where the bead of sealant had not made contact with the back baffle.  This leaves a direct path for the air/fuel to flow out of the tank.  In this picture you can see two spots where the sealant did not make contact with the baffle (up in this photo).  It looks like little dark shaped lines just above the bead of sealant.


Another area that needed work.  Again look for the dark line on top of the bead of sealant toward the right side of the picture.

I used my Solidworks skills, limited as they are, to design the circular patch.  The circle is 6.375" in diameter with 24 holes evenly spaced around the inner perimeter.  In this picture you can see the CNC as it starts to cut the outer ring.  The holes are already punched in this picture.  I used .032 2024 alclad for this step which is the same material/thickness that the back baffle is made of.
Next up I had to fit the patch panel to the baffle.  You can see where I measured and marked the center of the holes so that I could use those lines to center the patch plate.  In this picture I had just drilled the first upper hole that I used as a starting point for aligning the patch panel.

  Once I had the panel aligned I drilled the bottom hole, cleco'd the panel in place, and finally I match drilled the remainder of the holes.


Here is the final result of the match drilling.  Next up was deburring all the holes, inside and out, followed by a thorough vacuuming and wipe down with lacquer thinner.

Once everything was clean and deburred I mixed up a batch of sealant and proceeded to create a nice bead of it along the leaking seam using my 10 cc syringe method.  This is a picture of one of those beads after I had used my finger as a squeegee to smooth the bead and work the sealant into the cracks.

Next up was to run two beads of sealant around the opening in the baffle; one bead between the rivet holes and the inner 5" hole and one bead outside of the rivet holes.  The sealant you see oozing out is from that outer ring of sealant.  Finally I put a cleco in each hole to help pull the patch panel down to the baffle.
And finally I used some AD-41H rivets to fasten the cover plates to the baffle.  The rivets are twirled in sealant, inserted and pulled, then I put a smear of sealant over the back of the rivet just in case there was a rupture in the sealed portion of the rivet.

A closeup of one of the covers after riveting and sealing.

Sunday, March 18, 2018

Fuel tank leak tests

This week the Proseal had set up enough on both tanks that it was leak test time.  I set up my manometer test rig so that I could put 1psi of pressure on the inside of the tank.  There is a poor picture of it below.  I tested the right fuel tank first and it failed.....grumble, grumble, grumble.  As I mentioned in a previous post the back baffle is the hardest because you have to blindly rely on the baffle "pushing" a bead of Proseal into a nice filet.  The plans say not to use more than 3/16" bead of Proseal (diameter) and to put the bead just below the holes.  Well that's what I did...I thought... but it seems I must have put the bead a little too far below the holes which means the Proseal never made contact with the baffle.  Now I have to cut a 5" hole in that bay of the tank, reach in, add more Proseal, and then rivet a cover plate over the hole.  Its not a hard task and it's amazing how many people have had to do this exact same repair.

The left fuel tank is still attached to the manometer and doing well.  Once I have determined that there are no large leaks (like the one in the right tank) I will add about 5 gallons of 100LL fuel to the tank and let it sit in various positions.  100LL fuel has a blue dye to it and because of its viscous nature it will find holes that water will not.  So, I will let it sit a day or so in each position and then look for blue dye seeping from rivet holes or other sealed locations.

I brought both tanks into the house for a few days where its much warmer so that the sealant would cure faster.

Here is that horrible picture of the testing manometer.  You can see the plastic tube that gets hooked up to the vent fitting.  That tube drapes down to the floor and then back up (taped to a broom handle to keep it relatively rigid).  I put about 45" worth of water into the tubing and then use the air stem attached to the drain flange (on top as the fuel tank is stilling bottom up) to put enough air into the tank that the water level rises.  27" of water difference between the two sides of the loop is equivalent to 1 psi of pressure in the tank.  Then I watch the water level to see if there are any leaks.  Initially the water level started to drop indicating a leak so I sprayed the tank down with soapy water.  No leaks were indicated so I checked the manometer again and the water level had stabilized.  Best guess is that the tank is still cooling down from being inside the house.  Several hours later the water level had remained stable.

Oh yes, in the mean time I finally got around to cleaning up the edges of the bottom wing skins and clecoing them in place for final drilling.

Monday, March 12, 2018

Fuselage is in the house!

Big week this post!  Amy took a little vacation with her mother so I decided to take the week off myself.  I have lots to do and this seemed like a good time to get some of it done.

Leading off the week was my trip to Oregon to pick up my fuselage kit.  Its here but still not inventoried since I don't really have a place to store all the loose parts.  The kit comes in small "packages" of parts that once unwrapped need to be stored.  I have 30 days to do the inventory and I hope thats enough.  After 30 days I guess its on my own dime if a part is missing.

I also started the annual inspection on the RV-9 this week.  I changed out my spark plugs, did the compression tests, and pretty much finished the engine "group".  I also replaced my brake pads on both wheels while I was doing the undercarriage.  There was still a lot of pad left but I had ordered the parts so I figured it was a good time to do it.  I also finished most of the cabin group and that included finishing the heated seat wiring as well as machining a shim that took up a little of the design slack in the aileron linkage.  Overall things are looking good but I still need to finish the tunnel as well as the wings.

Finally, I sealed up the left fuel tank.  The rear baffle is in and now I have to let it sit for a week or two so the Proseal can cure.  Then its leak test time.

This is the storage location for the cabin cover.  It took far too much time to rig up a safe way to hoist that up and then secure it.

Left fuel tank with vent line installed. 


Left fuel tank with rear baffle in place.  Still no rivets but there is a cleco in every hole!

Same as the picture above but from a side view.

Sunday, March 4, 2018

Left fuel tank - almost done

This week I almost finished sealing the left fuel tank.  I kinda ran out of steam (energy) today before I got the back baffle installed.  I got all the ribs, vent lines, and related details completed but I was tired of breathing Proseal and MEK through a respirator so I called it a day.

Stiffeners riveted and sealed on the left fuel tank.
Here is the tank almost ready for the back baffle to be dropped in.  The fuel level sender still needs to be installed and a few shop heads still need to be encapsulated with sealant before the final baffle install.