When things go not quite right - trial by sand (1-pc fuselage development)
This is just a story of work in progress, for your entertainment. I certainly haven't been entertained by it, but perhaps you'll see it differently!
Rules of engagement - If you have something in a similar vein or artery to share, share it!
As many people here know, I've been working on a 1-piece mold. The plug took a lot loooonger to get ready then I expected. Many coats of sandable primer, many sessions of wet sanding, etc. It started from pulled pieces from our existing pod mold. Now that mold was made from a plug that was sanded foam coated in epoxy. For anyone who has done that, you know it is hard to get it very accurate. So there was a lot of reshaping and contour gauge work required. But finally it was looking alright. Some of you saw it at The Bruce.
A major step remained though. I intend the finished fuselages to have hardpoints and threaded inserts molded in place. Now to do that there have to be accurate alignment pins in the mold. To do that, there have to be accurately aligned holes for the pins in the plug.
No problem... I have a Bridgeport vertical mill. For those not familiar with old American Iron milling machines, they make one heck of a nice drill press!
So the first step was to find something of the right size to drill accurate holes for the pins. Actually, that was the second step - I needed the pins!
Of course, the pin size for 6-32 thread turns out to be something oddball. As in, I couldn't find anything stock that would work. With some searching and creativity, we found some nails that had the right shank size. Off with their points, a bit of rounding, and two pins were produced.
Now to drill the matching holes. Searching through the tools revealed that a standard size of a center drill would yield the right size hole. I thought that was particularly nice as center drills are very stubby which helps with accuracy. They are also made to high tolerances.
So I take the fuselage and - gently, wrapped in paper towels - clamp it up in a 6" Kurt vice on the milling table. Now the vice is aligned to the table, but how to align the fuselage to the vice? There isn't a straight surface anywhere!
I use the center wing section tool - the same one used to mold the wing mating surface on the plug - to align the fuselage to level. Level in roll, that is. For pitch, I eyeballed it. That left yaw. No good solutions there.
So I put a straight quarter inch rod in a collet chuck and use that as a probe. I go to the forward part of the pod, and measure the left and right sides by contact. Then go back along the tailboom, and measure the left and right sides. Then with a little math, and compensating for the thickness of the probe, I compute out the X and Y coordinates for the holes.
I really wish my mill had DROs (Digital Readouts). Without DROs, it ends up being turn counting on crank handles, and reading the scales. Easy to make errors or lose track due to inattention.
So then I drill the holes. That part was uneventful. And a few hours after I started for this "simple" job. But the holes were dead on the money, as best I could tell.
Now time to work some more on the parting board. I want something between the top and bottom half of the mold to ensure accurate alignment. Ball bearings. Large ball bearings. That's just the ticket. Or so I thought. 1/2" diameter.
I have a half inch ball end mill. One would think that would make the perfect hole to seat a ball bearing...
We decide to do this work on a drill press. NOT the right answer. That made the job trial and error for getting the hole depth so that the ball was bisected by the parting plane. I needed some sort of depth gauge for the job. We made a stack of metal scraps as a sort of cheap gauge blocks that did the trick. If it would slide over the ball, just touching, then the ball was the right depth. Turns out more error than I expected. If one didn't apply pressure, the ball would sit at the top edge of the hole rather than going in. So it was very easy to drill the holes too deep without realizing it. So easy in fact, that I just gave up and drilled them ALL too deep.
Now why couldn't one just set a depth and drill to that? The parting plane is curved to match the fuselage... And mounted on an aluminum bar (sort of mounted)... There is nothing super flat anywhere to use.
To make matters more fun, the board has some spring to it between the mount locations! So drilling exerts some pressure, and the material is not uniformly resistant... You might guess where I'm going with this...
No drilling, then it grabs and takes a bite. So getting accurate depth control was impossible. For that matter, with a two flute cutter, getting a round hole in the springiest areas was pretty much impossible.
So I just overdrilled them and pressed the balls in until they were the right depth.
A little CA along the edge to bond them in place - after all the certainly DON'T fit the holes! All is relatively good. Except for the one hole which had so much spring that the cutter practically exploded a cavity in the parting plane rather than cutting a nice hole. I used epoxy on that one - as a filler.
Now this mold is intended to bolt together, so a bolt every 3 inches... A bunch more holes drilled for pegs. Put wood pegs in holes, slide brass tubing sections (2" long and a lot of them) over the pegs.
But before putting in the pegs I had to get the plug bonded into the parting board. Modeling clay was used. The board had a gap around the plug so some balsa shims were used to lock it into place carefully and in correct alignment. That took a couple of hours. Then on to rolling out little worms of modeling clay and forcing them into all the gap. The clay was left below surface. The idea is that once one side is made, the plug is pulled and sanded flush. Then the plug reinstalled and the second side made. That should give a very sharp parting line.
Now another coat of wax, or two or three, was applied. Gotta be well waxed.
I didn't have the correct wax on hand, so I improvised. I tested various waxes I had on hand and found one that seemed compatible with PVA. At least it tested out fine on Mylars.
Put some PVA in a plastic cup, wet out foam brush with water, make foamy mixture, and brush on. Well, it wet out the parting plane just fine but it wouldn't stay everywhere on the plug. I guess the wax didn't do as well over the polyurethane. I had to wipe it all off. Second try didn't work any better. So the parting plane is PVA'd but not the plug. Now I'm trusting in the wax, which usually is ok, but...
In go the pegs. On goes the brass tubes.
Tooling coat time. System 2000, plus some graphite, plus lots of cabosil. When I'm done, it is like toothpaste. Black toothpaste. I scoop it on, mash it around, and stipple it in with a stiff bristle brush. I think it was 722g of tooling coat total. That's epoxy, not counting additives. I made a fair dent in a large container of cabosil.
Flash it with a torch. This stuff is just too thick, so I flash it a little more. As I've been told, it gets a little strange if you hit it a second time with a torch. Then I pop some bubbles with a pin.
Actually I think toothpaste probably flows better than this splooge.
Give it enough hours, and hit it with a second tooling coat. But that is after putting the sides on the mold. I just used sheets of plastic lined insulating foam, sprayed liberally with 3M77. A little tape on the corners just to make sure.
Now it is serious fill time. The fill is sand. Lots of sand, with some epoxy. Think mortar for bricklaying. Or thicker.
But about that sand. You want washed dried sand. After sufficient searching, and advice from Gavin, I found washed sand. Obviously washed, as it was still wet. Good sandcastle material.
My friend Alan works as a tech ed teacher - and has a kiln. Somehow he was foolish enough to volunteer to dry 50# of sand... It took him a week I think. But it is super dry!
So make up epoxy in 381g batches, mix in sand until you can't mix in any more, then trowel it into place. I used a stick to sort of beat it into submission, and to level the top. I put in one layer at about an inch thickness, waited a few hours, then put on the rest. I was quite tired of mixing this stuff by the time it was done.
Once it hardened on top, I peeled away the foam and started disassembling. First the heavy aluminum bar comes off the bottom. It was only put on with some foam tape anyway. Then unscrew a couple of support blocks, then rip off the parting plane. Rip is the operative word. It did come off, but a few of the pegs slightly bonded to a few of the tubes, so a little force was required at some places. Of course the parting board was made from something structurally weak...
It all came off though and the mold surface looked pretty good. Beautiful hemispherical polished holes where the ball bearings were. Plus a plug in the middle and some modeling clay. Pull off the clay, and pull out the plug. It doesn't want to come out. I stuck a metal rod full length into the plug and used it as a lever to get the plug out. Eventually I won that battle. At least the plug came out, and it didn't break, which surprised me. Even though the plug was heavy walled, I used a LOT of force. But it left a lot of paint behind.
Turns out the flashing probably got the epoxy hot enough to penetrate the wax a little...
And there is more bad news. The tooling paste was too thick so it left a few areas along the edge as voids.
Between the paint and the voids, the surface of the mold looked beautiful. Black and essentially flawless.
I used some tape to pull up some of the paint, leaving it at the edges as it would be needed when making the second half of the mold. Some paint came off, some I could scrape off with a fingernail. Some just wouldn't come off. I left that for later.
Now that the epoxy was cured I knew that just adding epoxy fill to the edges of the mold would be a long term recipe for disaster as it would break out. So I used a dremel with routing attachment and a cutter to mill out square the areas that needed repairing. Then I went in with a bit and undercut diagonally to give them a tongue anchoring them into the mold. This would give a solid mechanical bond.
More tooling coat, this time a BIT thinner! The tongues got milled fiberglass and the tooling got cabosil and graphite.
That got rid of the issues, mostly. On the back half of the mold, the repairs were invisible. On the front part, there was a little step that one could feel and see if one looked closely enough. That needed to be sanded down. I hate sanding in the mold but there wasn't a choice. So I sanded them out.
Now when the plug goes in, it leaves an almost paper thick (thin papar) gap at the sanded regions.
Oh, I forgot to mention - I block sanded the mold surface so the parting line really is nice and sharp.
About the plug going in... It sort of does and doesn't. I can't get it to seat all the way back in. It seems that the weight of all the sand and epoxy caused the mold to sag a little. Now the plug rocks in the mold. Although it does take some force to move it. It won't stay in place.
Now the task is refinishing the plug and starting again.
Last edited by G_T; Oct 13, 2010 at 12:58 AM.
For example, DRO Pros has a 2-axis mill kit for $599.
You can even build your own, with open source software.
Yes, at some point I'll put DRO on my Bridgeport. It has always been part of the plan. I searched around for about a year to find a J-head Bridgeport that had chrome ways and little wear, and got it for 1500. It needed a lot of cleaning - I pulled what must have been 40# of scrap out of the inside of the knee for instance. I added accordion way covers so that will never happen again. I've rebuilt the bottom but I still need to replace the bearings and bushings in the upper head. I expected to have to do that and purchased what I need at the same time I got the mill. It works quite well but has a little pulley chatter at top speed, hence the bearings and bushings. No big deal - just a long weekend of cleaning and rebuilding. I just need that weekend! Everything else works quite well.
The mill only comes into occasional use for DLG stuff, like when I milled a test mold for T-blades.
I'm not sure what this one weighs, but I think it is around 1700 pounds or so, and I set it up with true three phase power. Anything I've cut so far, it hasn't even noticed it was cutting. Once you get away from the tabletop units, mills become quite solid beasts!
But it is definitely NOT CNC. Besides, my minimum desired size for CNC is a cutting area of, oh, perhaps 10" by 40". That should do fine for DLG wing mold halves! I do not have the space or money for such a machine. I do have some very nice tools, but I've gotten them CHEAP! Except for things like collets, where I do not skimp.
Walt, the last mold came out fine. This one I did a bunch of new things on, and I knew I was taking a calculated risk not using PVA. I didn't anticipate the mold warping due to its own weight and that was just a stupid oversight on my part.
Ben, the plug was removed because of a technique I was using that gives razor sharp parting lines. When the plug is set in the parting board, there is a gap left. That gap is filled with modeling clay, but to a little below the level of the parting plane. Then after the mold is poured, the plug is removed. Now there is a ridge above the level of the parting plane. That is block sanded back, resulting in a nice sharp edge. Then the plug is reinstalled, and the second half of the mold poured (skipping steps in this description, of course). Really I probably could have handled everything that went wrong except the warping of the mold. There is no fixing that. At least I have enough epoxy and sand left for another full attempt. Time permitting, I'll start on the refinishing of the plug late tonight. Probably around middle of next week I'll be framing it all up for the second pour.
My little bit of spare time is getting split between mold work, aero design, and contests.
PS - Dave, I've already decided I'm not doing it this way again! Next time will be CNC milled. If I were currently up to speed on CAD with Rhino, next time would be this time! I admit to being very tempted... I just wanted to get the next wing design completed first. But in the short term, Phil and I need some more fuselages so hence this project. And since this one is optimized for an Edge wing I thought others might be interested in some fuses. Though it will take other wings of course.
Last edited by G_T; Sep 14, 2010 at 12:17 PM.
Perhaps I can come up with a few. You asked just in time though. Tonight I was going to strip off the polyurethane outer layer on the plug. It won't look so messy after that. The touch-up work should be relatively minor. In the process of making the mold I did find that there is an area that I mis-shaped slightly, causing a negative release angle. It wouldn't prevent pulling fuses but it would result in wear over time. I'll correct that, now that I know where it is.
Of course, the mould looks like @#$$# in some areas, with all the paint still sticking to it.
I didn't mention one of the other things that went on. In my attempts at paint removal, I tried a dremel with a plastic brush. I knew the brush was softer than the mold surface so it wouldn't do any harm. At least not abrasively. And that turned out to be true. It also would only barely remove paint. That is until friction heating got up there, at which point it removed paint and slightly marred the mold surface!
Some days you just can't win!
PS - I guess someday I should get a digital camera of my own. I used to do a lot of photography and printing, but much of it was B&W large format with big sheets of film in holders and big old cameras with bellows. Going out to shoot was sometimes like making a minor expedition!
Good luck Gerald on your endeavor. I am sure you will succeed in your endeavor.
|Category||Thread||Thread Starter||Forum||Replies||Last Post|
|Discussion||Not Quite Dead....All Things NQD||dJC_0175||Racing Boats - Electric||1||Jun 21, 2010 11:58 PM|
|Discussion||When things go wrong? When things go right?||Kimber||The Builders Workshop||1||Apr 10, 2010 05:54 AM|
|Help!||Frustrating weekend (again). FPV setup still not quite right||jbpilot||FPV Talk||9||Apr 05, 2010 11:49 PM|
|Discussion||Warming...the models are right, except when they are not.||MtnGoat||Life, The Universe, and Politics||143||Oct 05, 2009 08:33 AM|
|When a 2 cell lithium is not quite enough and a 3 cell is too much...||Xnaron||Batteries and Chargers||16||Sep 22, 2003 09:53 AM|