Confession: the hours logged here are pretty much a wild guess. I think I’ve had five or six nights in a row where I was just about to fall asleep and remembered I hadn’t written anything up about the build. Which is to say, I’ve been doing stuff, but not really keeping up with the build log.

When we left off last time, I was about to start worrying about the latch setup for the oil door. As previously mentioned, the approach here will be to use regular old piano hinge, with a bowden cable to release them. The first step was to plan how I’d lay out the hinge halves. At first I got really hung up on wanting the edges of the eyelets on the door to line up with the edges. That was a problem because the door opening is about 1/4” narrower than the eyelets I’d want to line up. I started to think of crazy ways to handle that, before realizing that it was unnecessary.

I started by laying out the hinge on the cowl side. I cut a piece of hinge long enough to cover the opening, plus have an extra eyelet on the forward side, to act as a guide for the cable. I also spent a lot of time trying to think through how to ensure the hinge on the door would clear the cowl, and finally I decided to just line the cowl side of the hinge up with the opening and call that sufficient. That meant I could just go ahead and attach the hinge to the cowl without worrying about fitting everything beforehand.

So I laid out the rivet holes, drilled the hinge and cowl together, and got everything prepped. As usual for metal-fiberglass bonds, I mixed up some loose flox and painted the joint with it before riveting things together. Another consideration here is that due to the recess of the door flange, the part of the hinge that projects forward sits a good bit above the cowl interior. To handle that, after I’d painted on the loose flox, I thickened the mixture up and put a big blob under that forward projection, to fill that gap. The next morning, it had cured up nicely:

Next up was the hinge half on the door itself. The big consideration here was shining the hinge off of the door to get it to mesh nicely with the other half. My initial measurements indicated that some .063” aluminum would work nicely, so I cut a shim off some scrap and went ahead and riveted it to the hinge I’d cut. Things seemed to line up nicely, and I drilled the rivet holes to attach that stack to the door.

But once I was testing the door movement, the hinges wouldn’t mesh any more. Seems that I needed a little more shim action – so I ended up cutting another shim out of .032” scrap (I think), and I got to drill out the rivets in my hinge-shim stack and add the extra shim. With that in place, things lined up very nicely. After testing the door action a bunch more times, I decided it was time to permanently attach everything else – not just the latch hinge half, but also the actual hinge pieces, to the cowl and door. I wanted to do all these at once so as not to waste epoxy – even one pump of epoxy yields about five times more flox than is needed for just one of these joints.

I did take a different approach to these joints. Previously, for these metal-glass joints, I painted on the flox, clecoed things together, and riveted immediately. This was invariably a messy experience; this time, I decided to try just checking things, letting the flox cure, and then come back to do the riveting later. This worked out pretty well – though for some reason I didn’t take any pictures of any of this work…

After going back and riveting, it was time to put everything together and test the fit and action of the door. Things worked pretty well, but the one thing I noticed that bothered me was that the door didn’t quite sit flush with the cowl once I had the spring installed in the hidden hinge. It seemed that that spring – which is there to make the door pop open when it’s unlatched – was pushing up on the hinge side of the door just enough to keep it from sitting flush.

I was able to close the gap a bit simply by tweaking the hinge itself, but it still didn’t really get the job done. I suspected that the flexibility of the door itself was contributing here, and I spent a fair amount of time trying to think of ways to add some stiffeners to the inside of the door, but none of the ideas I had seemed like they’d be especially pleasant to implement.

Then I read an absurdly simple solution on another build log, where the guy simply used the piece cut out of the cowl, epoxied to the back of the door, to add stiffness. Since that piece, by definition, already has the same curve as the door it ought to work well. And here was a case where my poor shop cleaning habits worked in my favor – I still had the original piece I’d cut out of my cowl sitting on the work bench.

It was still plenty big enough for this job, so I got to work. The big fun here was to do the necessary trimming to get it to fit inside the door. I went a little above and beyond by having it extend around the edges of the hinge; probably unnecessary, but since that was the exact area where the door was popping up, I wanted to add rigidity as far in that direction as possible. So instead of just cutting a nice square-ish piece, I had to carefully fit a cutout around the hinge.

Here’s a look at the trimmed cowl piece after trimming. I’d already added some masking tape in preparation for the impending application of epoxy:

Next, I drilled the two pieces together so I could use clecos to hold them while epoxying, did a whole lot of work to rough up the mating surfaces, and also added packing tape on the cowl flanges. The last thing I wanted to do was accidentally epoxy the door to the cowl, after all.

Then it was time to make another batch of flox, paint the surfaces again, cleco everything together, and clean up all the squeeze-out. After that, I reinstalled the door on the cowl, making sure to use both the hinge pins as well as the cleco holes I drilled on the perimeter, to hold the entire assembly in the exact position I wanted it, and thus ensure that it cured with the desired shape.

Here’s a look at the assembly just before I added the perimeter clecos; the packing tape on the flange is clearly visible:

The next morning (this morning, actually) was the age-old fiberglass routine: go find out if I accidentally glued two things together that weren’t supposed to be. It’s always a bit of a relief when the parts that are supposed to separate do just that.

I revisited the setup after work; the big question mark was seeing if this stiffening actually helped the fit problem I was trying to solve, or if I’d just wasted time. I’m happy to report that it now fits up very nicely. I may still sand down the top forward corner (lower right in the photo below) just to really fine-tune things, but it looks pretty good as-is:

I also spent some time sanding off excess epoxy inside the door; it seemed like a lot more squeezed out while this was curing. I tried to make a nice semi-rounded transition between the door and the reinforcement, mainly so there aren’t annoying sharp edges here, but also just because it looks nicer. And finally, it was time for one more function check of the door. It’s a little rough popping up when the latch pin is pulled, but I think I can fine-tune that just by sanding the edges of the eyelets where they contact. For the most part, though, the thing pops up nicely:

I told someone this weekend that when the oil door was done, I was probably done with fiberglass for a while. Technically, I do still need to take care of glassing in the cooling ramps on the upper cowl, and it’s looking like I am still going to need to add a blister to clear the mixture arm, but I think I’m about ready to switch gears for a while. Plus, it’s supposed to cool off later this week, and low temperatures don’t really agree with fiberglass – if it’s cold, the epoxy won’t kick off and cure properly.

So I’m going to pull the engine baffle kit out of the bin and get back to work on that. I think it’ll be nice to get back to doing some regular old sheetmetal work.

Yes, shenanigans. It’s the most appropriate word, I think.

The overall goal was to apply filler around the oil door opening so I could get a nice paint gap as well as have the door be flush with the cowl around it. The first step to this was to basically sand off the factory-molded lip around the oil door flange, so I could build it back up. Then there was the really fun part – figuring out a way to add some material to the door itself so that when I laid up the micro, I’d get a reasonably usable starting point for the gap.

First, I tried just using layers of electrical tape, kind of like I did for the cowl pin covers. Problem was, I wanted more layers than I used for the covers, and trying to lay on a bunch of tape while still keeping the edges tight, and invariably, after about three layers I’d get a wrinkle. Then I made a brief attempt to use self-fusing tape, which is thicker, but that just didn’t go well at all.

Finally, I tried using some edge molding I had laying around. This is stuff you might stick inside a lightening hole to avoid chafing, and I thought I could apply it around the door. At first, it didn’t really want to lie flat, but with some strategic electrical tape, it seemed OK:

I followed this up with another layer of electrical tape around the perimeter, and checked the fit against the sanded-out door cutout:

Then it was time to glop on the micro, reinstall the door to squeeze out the excess, and build up as needed around the perimeter:

This ended up being way more micro than I needed, so the next day I had a whole lot of sanding to do to get back down to the proper contour. As I worked that down, I could see that the edge molding apparently hadn’t been flush all around – there was one particular spot where it must have popped out a bit, and so I had an excessive gap. I went ahead and sanded everything down, then worked on fine-tuning the gap, hoping I could make it less obvious, but in the end I wasn’t happy with the result. The annoying spot is at the bottom right of this photo:

So…I sanded all that micro off and started over again. This time I made another attempt at just wrapping with electrical tape, but by being VERY careful I was able to get the whole thing done with no wrinkles. Once again, I glopped on the micro, added the door, and built up as needed:

This time, the results were better. It was a little annoying to go through thee exact same contour sanding operation again (at least I used less micro this time), but after it was all said and done, things were looking pretty decent, though I think I’ll still end up doing more fine-tuning of the gap here when I finish the cowl as a whole:

So now I suppose it’s back to the hinge and latch…

I have been working on the project – maybe not as much as I like – I just haven’t been writing about it. Picking up from where I left off last time, I got in my Aircraft Spruce shipment, and got right to work on the oil door.

To start with, I located the door in the cutout and drilled a few holes around the perimeter; this allows me to cleco the door in place while I work on other fitting-related items. First up in that department was the hidden hinge. This is pretty straightforward; the measurements given for locating the hinge half on the cowl looked good, so I drilled and checked it in place.

The half for the door required a bit more work, though. It didn’t want to sit flush against the door itself; I suspect this is due to the cowl having a bit of a conical profile here, rather than a pure cylinder. Other builders I’ve read have used shims or flox to fill the gap; I instead decided to tweak the bends on the hinge itself to get a good flush fit. Basically, I needed to close up the sort of C-profile a bit on one end to pull the flange down. Then the entire flange needed to be bent a little to get a flush fit.

After numerous iterations of this bending, things were nice and flush and so I went ahead and drilled the hinge to the door as well:

Once this is riveted in place, there will be no visible hardware for the hinge at all:

The next logical course of action would have been to work on adding filler around the door to set the paint gap, but at some point I unilaterally decided I needed a break from fiberglass, so I decided to poke around at other firewall-forward things. This led me to revisiting the throttle and mixture cable routing. Specifically, I wanted to see if I could tweak this setup to help with my potential clearance issue between the mixture rm and the lower cowl. I really don’t want to have to mold a blister down there for clearance if I don’t have to.

To make a long story short, I determined that I could position the mixture arm further back, but not with the cable mounts as provided with the Showplaces cable bracket I bought. But all I really needed was to move the cable down and sideways a bit, and add a bit of angle. Normally I’d fabricate a bracket for something like this, but the bracketry here is 4130 steel instead of aluminum, which I imagine is related to this stuff being really close to hot exhaust pipes.

So I ended up dusting off some 20-year-old neurons, relearning how to do basic CAD modeling, and designing a slightly modified cable bracket, which I then ordered from SendCutSend. I’m pretty optimistic that this will help a lot with my cowl clearance issues:

From here, I guess I need to get over my anti-fiberglass feeling and get back to work on that oil door. I think that’s the last thing I really need to do before I can move on to the engine baffles – which is the task that started me on this whole cowl odyssey…

We rejoin the fun times from Sunday morning, where I went straight to work on all that micro I laid up around the pin covers. Unfortunately, my electrical tape trick didn’t work out perfectly – the tape ended up sliding off the edge of the pin covers in a couple spots. As a result I had no gap at all towards the aft end of both covers. The plus side is that since micro is way easier to sand than epoxy and glass, this wasn’t a huge problem. After the initial contour sanding, I went to work slipping sandpaper around the cover edges to clean up and open up the gaps as needed.

The resulting gaps aren’t what I’d call “perfect,” but this time I really am satisfied leaving it alone:

Looking at it from an oblique angle, the covers line up much more nicely with the cowl curve thanks to the contour sanding, though again the aft end of the right side cover is a little proud…oh well:

The next task will be the oil door. I pulled the upper cowl again, but before getting to the door, I finally took the time to sand the rivet line at the firewall edge. This is harder work than messing with the covers, but at least it’s not finicky. Basically I just go to town with 40-grit until the area is relatively level, then follow up with 100 grit on a soft block to smooth out the scratches.

I also went ahead and did a first round of scuffing on the perimeter of the oil door opening. The finish here was kind of crappy to begin with, especially the edges of the recess for the door, so I fully intend to reform this whole area with filler as part of the oil door fitting process. The micro needs a rough surface to adhere to, plus I want to sand the existing recess edges back to a bit of a bevel, kind of like I did with the pin cover openings. I’ll still want to do more work on that bevel, but this is a start:

For Monday’s work session, I did some more detail work. While I’d refined the paint gaps along the sides of the cowl and the cheeks, the spinner joint was still way too tight. So I joined the halves on the bench and spent about half an hour working on those until I had a nice gap in place:

Then it was on to the door itself. The door is provided as a prefab fiberglass unit, but it’s oversize and needs to be trimmed down. There are scribe lines on the door as trim guides, but upon close inspection they’re not perfectly straight, nor are they even. So while I used them as rough guides, in the end I put in my own work to properly squares up the trimmed door. Here’s the door before trimming:

My approach was to first rough cut one of the curved sides, sand it roughly straight, and then fine-tune it. For the fine-tuning, I glued sandpaper to a piece of aluminum angle; resting the door and the angle on the table made it easy to gently sand a very straight and very square edge.

One that was done, I used a square to lay out the other three cut lines, and repeated the process of rough cutting and then sanding to exact shape. Finally, I traced out radii for the rounded corners I wanted, and hand sanded those to shape. My cuts ended up a bit outside the scribe lines in some spots, so I also hit the perimeter of the door with 100-grit on a soft block to smooth things out a bit.

Here’s the trimmed door sitting in the cowl cutout:

The next question will be how I want to locate and fix the door in place on the cowl while I do the hinge and latch fitment. I might just end up drilling a few perimeter holes where I can cleco it in place, but I’ll want to think on that a bit. Food for thought while I wait for my parts from Spruce to come in.

Didn’t put in a ton of shop time today, at least partially because I’m annoyed at it still being 90+° in October. You’d think I’d be somewhat used to that after 11 Houton summers, but nope, the way it stays hot so late still bugs me every year…

Anyway, I decided to work on finalizing the cowl pins. I knew I was going to need some sort of bend at the forward ends, but I’d been thinking of a few ideas for a day or so. I did some poking around with other build sites, but didn’t really find any solid information. So today, I found some scrap Romex lying around and found that it was a good size for doing some tinkering with what shape I wanted. I settled pretty quickly on having the pin bend inboard about 30° at the forward most hinge eye, then bending a little eye shape about an inch and a half away. After mocking it with Rolex, next I did a trial bend on some extra pin material I had lying around, just so I could do a trial of actually inserting/removing the pin.

That worked out well, so I moved on to the real pins. Nothing really exciting here, though it did take a bit of trial and error to get the bend just right. I wanted the eye to be clear of the back of the pin cover, but only just barely, so I could still easily grab it with needle nose pliers. Then it took a couple iterations to trim each pin to the proper length and grind a point on the aft end to help it get through the hinge eyes.

The resulting business end:

With the pin fully inserted, it tucks into place nicely:

With that done, I had the opportunity to get a look at the cowl sort of in its final form, with the pin covers installed. At least the gaps look just fine from 10’ away:

Speaking of those pin covers and they gaps…as I sort of anticipated yesterday, the more I look at those gaps, the more they annoy me. In the midst of thinking about that today, I randomly tried adding washers behind the pin covers, and found that having them sit just a bit higher does help aesthetically. So now I’m considering adding a little material to the packing there to make this extra spacing permanent…but then there’s the question of correcting the gaps.

I’m considering opening up the edges a bit, then packing micro back there and installing the covers over it to set better gaps (and, by extension, the nearby cowl contour). The main problem there is that I’d need to find some way to tape the cover edges to establish the gap, like maybe five or six layers of electrical tape. It might work, but it also feels kind of fiddly to me. So I dunno, I’ll still be mulling that over for a bit.

When we left off last time, I was dealing with some weird gap appearances around the pin covers. First step for dealing with this was to work on sanding the cowl counters to match the dummy pin covers. I put electrical tape on them to save them from sanding – not for cosmetic reasons, since they’re throwaways, but really just because I only want to sand the cowl, not them – installed them, and went to town. At first I was trying to use a small block, but I was finding it really awkward. Instead, I decided to break out my oscillating multitool and use it with a 60-grit sanding pad. This worked out really nicely; I was able to carefully localize the sanding to just outside the covers, and ease the material off nicely. Doing this did leave a bit of a bug just outside where I was sanding, but the block I was using before worked well for cleaning up that contour.

With the dummy covers showing a nice consistent contour all around, I swapped in the real covers, expecting things to look much better, Well, they did look better, but still not like I was expecting. For example, the right-side cover still had a really prominent overhang going on in the upper rear area:

I find this mildly annoying. The idea of this pin cover kit is that you do all the fitting with the dummy covers, then swap in the real ones, which are slightly smaller and thus allow for a paint gap. The implication here would seem to be that all the contours should be the same, but my experience isn’t showing that.

Another reason for using the dummy covers is to avoid marring the real ones, but at this point I decided there wasn’t much to do but to instal the real covers and use those as a visual guide for refining the contours even more. As before, I covered them with electrical tape for protection, and went to town with the oscillating tool again. This time, of course, I was really careful working near the covers.

In the end, I got things to a point where I was satisfied with the overall fit. It’s not picture-perfect, but at some point I need to calm down here or I’m going to end up sanding this cowl on a daily basis until I die of old age.

Similarly, I’m still not 100% pleased with the fit when viewed from the side. Again, given how the dummy covers are supposed to be used for fitting, I’d expect to have a nice even gap all around, but instead things are definitely uneven. But I don’t really have any desire to go nuts here trying to make things perfect. At least not right now…who knows what dumb idea I might given a few days to mull over it

Anyway, here’s the side view:

Also shown above, but maybe less obvious, is some more fiberglass work I did here. When I originally opened up the cover backing to allow access for the pins, I took the opening all the way to the edge of the pin cover opening. The results, when the real covers were installed, were obvious openings at the forward/rear ends of the cover. I guess they might not have really been noticeable in the real world, but they were bugging me…so I decided to fill them in.

To do that, I did the usual scarf-joint dance: sanding the existing glass to a knife edge, then laying up some more small strips at the ends of the openings, with the dummy covers in place to keep things neat. Then, the following morning, I got to split the cowl halves again and do some more cleanup. Now those inner flanges extend a bit all around, and provide a nice backing for the covers:

Back to this evening, I finished things off by fitting the lower cowl back on the plane. I think my next move is to take the temporary pins I’ve been using and get them trimmed to length and formed for actual usage. The big question in my mind is what I want to do with the forward ends. These need to project past the forward extent of the hinges, and I need to be able to reach in with something like needle nose pliers to grab and extract them. Not sure if I want to just put a little 90° bend up here, or bend a little grab ring, or what. I think this calls for some research into what other people are doing.

OK, time to get these cowl halves split again. The portion forward of the covers was pretty easy; a hacksaw blade slips nicely into the gap, so I could cut right up to the cover. Then I went inside and used a Sharpie to mark a cut line from there back to the back edge of the cover. The good old oscillating saw let me make shallow cuts along those lines. Then I got a flat blade screwdriver into the cut and was able to drive out the covers, and finally finish the cut with the hacksaw blade as needed:

Next up is adding nutplates for those screw holes, then removing excess fiberglass. I need an opening so I can actually access the pins to split the cowl. I decided to just make this opening the entire length of the cover cutout – after all, I only really need the material to locate the nutplates, right? (more on this decision later…) After that trimming, plus the requisite cleanup with sandpaper and such, I rejoined the cowl halves for a fit check:

Then I figured I’d test-fit the final covers to see how things looked. These are a bit smaller than the dummy covers; this is intentional to leave a paint gap. Interestingly, the right cover (shown below) ended up not exactly centered in the cutout. Not entirely sure how this happened:

A different angle shows how far the cover is inset on the top side; this is expected, as I mentioned in yesterday’s post:

It’s kind of amusing the decision I made yesterday which was intended to avoid working with a bunch of micro here, because it looks like I’ll be doing that after all, thanks to that odd final fitment issue (the other side is similar, though not as bad). I think what I’ll do is first sand the area to match the contour and fix the inset issue, using the dummy covers again. I assume that this will yield an irregular gap, which I can then fix in the usual fiberglass fashion: stuff in a bunch of micro and sand, sand, sand. We’ll see.

Late-breaking update: I went back out and decided to try fitting the dummy covers in place. I had the thought that maybe somehow the screw holes and/or nutplates ended up slightly out of location, and that was what caused the real covers to look so off. If so, then when I popped the dummy covers back in place, I should see misalignment with the screw holes.

I did not, in fact, see any such misalignment, and I was able to screw the dummy covers in place just as easily as the real ones. So now I’m wondering if this is a visual effect from the covers not matching the cowl contour. I guess the upshot is that regardless, my first move is to get to work sanding those contours, and then I can decide from there what I want to do next.

Today was the day I finally did it – I convinced myself to stop overanalyzing and just go ahead and get the glass work done for these pin covers. I’ve been sort of putting this off for a while, probably because there’s a sort of finality to actually mixing up fiberglass – though the reality is that you can cut/sand off pretty much anything and redo it. I’d rather avoid that, though…

I did have one epiphany since the last work session, though. At that point I was wondering about dealing with the covers not quite sitting flush with the cowl (the curvature doesn’t match 100%), and I’d had the thought to have them sit a little proud and then add micro to get the contour right.

The epiphany was to basically do the opposite – set the covers so they were flush at the highest points, and then sand off the fiberglass to get the contour I want. In one sense, it’ll be tougher – sanding an actual layup is a lot more work than sanding micro – but on the other hand, I don’t have to deal with coating the front of the covers with release agent, slapping on a ton of micro, and then dealing with the inevitable voids. The other concern could be sanding all the way through the layup, but since I’m adding material on the inside to create the flange behind the covers, that’s not a real concern.

So I got to it today. First up were a couple of prep steps – the dummy covers used for fitting needed a few coats of wax to act as a release agent, and the insides of the cowl needed cleanup. Alternating between the two tasks worked pretty well – put a coat of wax on the covers, then go to town inside the cowl with the Dremel on the flox I added a while back. Once I was happy with that, I got the cowl assembled on the work table, and paused for a bit to allow the wax to dry.

Next was taping the dummy covers in place. The holes are snug enough to grip the covers just a bit, which means the tape isn’t doing a ton of work; it’s just there to help ensure the covers don’t move while the glass work is happening. This was also my opportunity to get the flush/proud/inset positioning I wanted. This photo of a cover taped in place highlights how the ends of the cover at relatively flush, but the middle is inset. The overhead lighting really accentuates the resulting shadow:

On the inside of the cowl, the flat backs of the covers don’t match the cowl really at all, so the first step is to mix up some micro and smear it around the edges of the cover; this provides a relatively even surface to top with glass plies later on. I was a little worried about moving the covers around while doing this, but they stayed in place very nicely:

While that micro was getting tacky, I moved on to the glass portion. I’d already measured before to determine the proper size for the glass pieces, so I used that to cut some glass and do the tried-and-true prepress technique – wetting out the glass plies between two layers of plastic, squeegeeing the excess off, and cutting out the pieces I needed. Then I just got to lean way inside the cowl – again – and stick them in place. Some work with gloved fingers and a chip brush worked out air bubbled, and then it was time to sit back and wait for it all to cure:

Next up will be removing the tape, drilling the screw holes, and then I get to lean inside the cowl yet again and re-separate the halves – at the moment they’re bonded together. I’m a little nervous to see how things look with the real covers in place, but I should get that chance tomorrow, I think. Another one of those moment-of-truth situations…

Let’s see – looks like our last installment was when I went and did some void filling around the prop spinner. I did work on sanding that the following day, even though I didn’t write about it then. And of course, in the process I saw quite a few more voids that I neglected to fill.

So when I picked up again last night, I mixed up yet another batch of micro, and went around applying it to yet another batch of voids. Then today was, of course, sanding day again. After working over everything, I did turn up a few more pinholes, but at this point I think they’re all too small to effectively fill with micro – I’ll worry about them whenever I do a general pinhole/finish skim coat on the cowl halves.

After that, I decided I’d fit things together sufficiently to get back to working on the pin cover cutouts. A few rounds of trial-and-error, slowly massaging the cutouts, got me to the place I wanted, where the dummy covers fit nicely into place. The only thing that annoys me is that they won’t sit flush against the cowl the whole way around. Whenever I lay up the flange behind these, I’m just going to have to decide how I want them to sit and deal with the consequences…though I’m a little tempted to let the cover sit a little proud, and do a little micro buildup around them to get things sitting a little nicer. On the other hand, that sounds like a lot of work for maybe limited benefit.

Anyway, the dummy covers fit nicely now:

I also worked on setting some more paint gaps, specifically around the outside cowl corners, and where the spinner halve meet. I’m pretty satisfied with those at this point.

I think my next move is to do the layup behind the pin covers and get that flange established where it needs to be. Which means I’ll have to make a decision about how the covers sit against the cowl. It also means I’ll have to stick my entire upper body into the cowl, just to reach that inside area where the flanges will go.

Overall, I think I’m getting perilously close to being able to resume the actual cowl work steps outlines in the construction manual. I think maybe the next things will be attaching the cooling ramps, and working out the oil access door…