N891JF Speedbrake

N891JF Speedbrake

created May 2014

The runway I fly out of is only 2600' long, and 40' wide. That sounds long enough, but KRs are very slippery airplanes. Because they are fairly aerodynamically clean and the wing is so close to the ground, once the plane gets a few feet or inches above the runway, it glides like a hockey puck. This prolongs the glide and burns up precious runway, and increases your susceptibily to cross-winds, gusts, and other nasty stuff before touchdown.

KRs need something to slow them down, either flaps or a speedbrake. Yes, I do know how to slip the crap out of this airplane, and that's helpful, but at some point you have to straighten the plane back out, and even if you do that perfectly at the point of touchdown, you still need something to slow you down until the brakes take effect.

The effectiveness of either flaps or speedbrake on a KR is well known. Flaps are a bit of an effort to retrofit, but a belly-mounted speedbrake is relatively easy and practically non-invasive. All work on the plane is done at the seat back level...pretty simple stuff.


I started this effort by drawing up the mechanics of it all in CAD so I could match the throw of my actuator to the desired deflected angle of the flap, which I wanted to be about 60 degrees or so. It turns out that the pivot point for the actuator needed to be about 5.25" aft of the front edge of the flap. I made the flap as long as I could, with about an inch of clearance between it and the runway to handle gear deflection. You can see the "new x-member", a piece of 5/8" x 5/8" spruce that I epoxied into place between the floor and the bottom of the aft spar, which served as the main attachment point of the piano hinge. I couldn't use the existing one, because it wasn't really joined to the bottom of the spar, although I could argue that it certainly isn't going anywhere.


Here's the flap itself, made entirely out of scrap material. The foam is 1/4" thick Last-o-foam, the outer layer of carbon fiber is some 5.85 ounce material left over from awing covering on N56ML, the plywood strip where the hinge will mount (top of photo) and at the center strip where the actuator will mount are from KR firewall leftover, and the tapered spruce trailing edge (at the bottom) is a strip sawed off the aft spar while contouring for the airfoil shape.


This was a layout (not wet yet) of another layer of inner reinforcing carbon fiber. The triangular piece is a litterally a scrap, so I did what I could to reinforce the center section and spread the actuator's load. Hence the 45 degree angle of the weave. It also wraps around the other side. As yo ucan see by the epoxy on the scale, it was show time, at this point.


In another cup (not shown) I mixed in some micro to the point that it was still runny, but not so stiff it wouldn't pour onto the foam (probably 50% epoxy and micro by volume). Then I poured it out in this zig-zag pattern to make it quicker and easier to spread into the foam...which is how I've learned to do this over the years. As you can see from this photo, the inner wrapping of carbon fibre is unidirectional that's probably 9 ounce fabric...again, because that's a scrap I had left over from some other project.


I spread peel-ply over the whole mess and let it cure. Here's the difference in the final surface with peel-ply and without...it's huge. No pin holes in the covered surface, and after the cutting the flap out around the periphery on the bandsaw (this probably took two minutes at most) it was ready to paint...not that I'll ever get around to that!


Here's a cross-section of the nose, and the mounting of the piano hinge. Note small cross-section of the lightweight hinge. Nothing fancy...keeping it light. If you don't think it's strong enough, I once used a 3" long section of this stuff to do pullups in my basement to prove the point. It's up to the job, and it's the inexpensive "rolled" hinge, rather than continuous. The screws are number 6 wood screws. I pilot-holed a smaller diameter hole for the screw, and put a drop of T-88 in each before screwing. Between that and the 5-ply finnish plywood, they're not going anywhere! Looking at this cross-section, the ends need to be "heeled" by sanding out the foam and replacing with a fillet of flox. With 20/20 hindsight, what I should have done is spend the extra time to taper the sides of the flap like I did the trailing edge, so the top and bottom layers of carbon fiber would meet up and form a nicely tapered edge that would be bullet-proof. This will work fine though, and to tell you the truth, it's flying fine right now just like the picture above. I really should spend the time to fillet it and make it look nicer (and have stronger edges).


I used a "Motion Industries" actuator that was the darling of the RV crowd years ago when I built N56ML. It worked well in that plane to operate the huge 54" long flaps, so running this little 30" wide belly flap will be easy. At first I bought another much cheaper unit off of ebay with integral limit switches, but it was almost twice as heavy...mostly because of unnecessarily strong body, so I saved it for another job. These older Motion actuators have a reputation as having failing brushes, but for $18, you can get a new set from Motion and install them in a few minutes. I bought some for this installation, but only because the wires had fatigued where they exit from the plastic housing. Now I have new brushes AND it works too. I was fairly amazed that they still stocked parts for something that's not been in production for years, and the price was very nice! Regarding reliability, 2800 landings on N56ML, almost all with the huge flaps deployed, tells me they are pretty reliable, unless you try to rip the wires out of the motor and break one.


The Motion actuator, plus the bracket and bolt that mount it, was slightly more than 1.5 pounds, which is hard to beat. The chinese ebay actuator was twice that. Looking at this picture makes me realize that I missed another opportunity to lighten the bracket on the mounting side. I'll be smarter on my next airplane!


Location of the actuator bracket was dictated by location of supporting structure and dodging the elevator pullies, although N891JF has 1/4" thick plywood seat backs, so it could be argued that it could be mounted directly do the plywood.


I didn't think anybody would ever see it, so I used these two spruce scraps to space the bracket out to match the existing vertical and give me more width to mount to. At this point they were barely held in place. I later marked them and covered them with another piece of 1/4" finnish plywood to take the high shear loads the actuator would dish out on flap deployment. You'll see the final look in the photo at the bottom.


I needed an aluminum tab to push on the flap, so again I drew it up in CAD so it have sufficient clearance for the clevis...and would look decent. You could wing it without this, of course, and save yourself a few minutes, unless you turned out something hideous and had to make another one or two...but this is pretty basic stuff. This piece isn't as big as it looks...the hole is 3/16" and the angle is 3/16" thick, More than enough for the job, but the next size down is 1/8", which I wasn't sure could handle it.


I then positioned it in the right place and match drilled two holes for the #10 mounting screws. The flap is full width of the fuselage bottom, and 9.5" tall. This gives about an inch of ground clearance when fully deployed to 60 degrees. So far, I haven't hit anything with it, but that may change!


I'd have used two #8 screws, except the "weld nuts" from Mcmaster Carr start at #10 size. You could use something else, but I figured these would come in handy somewhere else, so I ordered a bag of them (a lifetime supply). They are perfect for this job...that's for sure. I later covered this with a patch of carbon fiber, peel-plied, and it's hardly noticely on the bottom of the plane now.


The whole flap with hinge and actuator mounted is 832 grams, or 29 ounces, 1.86 pounds.


Then I screwed the piano hinge to the bottom of the plane, being careful to hit the 5/8" strip of spruce that I'd epoxied in place between the aft spar and the plane's 3/32" plywood bottom skin. I used simple #6 wood screws for this, given that there is little tension on them...mostly shear.


After mounting, I flipped the flap up and marked where the angle would need to go through the skin, and slotted it out with a Dremel tool, knowing it would have to be elongated to deal with the swing of the actuator. Here I'm holding the flap up with a drill bit inserted through the pin hole, so I can line up the actuator and place it accurately on the seat back before match drilling the bracket and bolting it in place.


This is the limit switch assembly, made up of pieces that were formerly used in other places. I spent way too much time trying to figure out how to stop the motor at both ends of the travel, and still enable moving it back the other way when the time came, since I'd already wired and installed the switch in the panel, and it was now inaccessible. Eventually it occured to me that a diode would make it work from this end, and indeed it does. Without the diode, the limit switch kills the power to the flap, and now you can't get it back down again! This whole assembly is 30 grams...right at an ounce.


This is essentially what is happening with the limit switch part of the assembly above (the green one). I apologize for the quality, but it was a sketch for me, not the public, when I drew it up in May. The little green nub protruding from the bottom of the left side of the limit switch is the button that the flap presses against to break the connection and stop the power from pulling the flap up anymore, although the actuator is installed so this is about as far as it would go anyway, except for some room in the clevis to adjust out freeplay as things wear a bit. The switch shown here isn't really accurate, because in reality you need double pole double throw...a momentary pole to go down, off in the middle, and it should stay up for powering the flap up on a go around (then the limit switch kills it at the top). I used an Otto 10 amp, #T7-231G1, sold by Aircraft Spruce as their 424081. I'll improve on this schematic to provide more details of the switch. I just did this one as an effort to divide it into the "UP" part, and drew another as the "down" side. The other switch shown above is just an indicator that the flap is "DOWN" (or at least, not parked UP)...a green LED on the top of the panel, right next to the red OIL light and yellow "iEFIS problem" light,which would warn of any number of other things, such as high oil temp, low oil pressure, high EGT, etc.


Here's the view from the tail. You could argue that the mounting block is way too big, but the intent was to provide a longer area for a different actuator (the ebay one) as an alternative if ever needed. As you can see, this is a pretty simple installation, and it works fine. All of this 1/4" plywood in the seat back is still bothersome to me...about four pounds of weight that doesn't need to be there. N56ML only has the two 5/8" x 5/8" verticals in the center for the glass/foam seat backs to lean against, but they do much, as the hinge at the bottom spar and the shelf up top take the majority of the load. Plywood is overkill, and unnecessary, but was "accepted practice" in the 90's. The plywood makes working (and stowing baggage) almost impossible behind the spar, so everything is done through these little removable panels that you see on the seat back. More weight. I have a feeling that the oscillating miracle tool will make quick work of this stuff someday!

Now that I have about fifty hours on the plane since adding the brake, and probably 100 landings, I can tell you that the speedbrake is a wonderful thing. It's a great way to slow the plane down while landing. The difference with it stowed and deployed is huge. Because the drag is behind the CG, it actually improves directional stability when deployed. And something that surprised me is that it lowers stall speed by about three mph, as unlikely as I'd have thought before actual testing! And if it slowed the plane down, I sure can't tell it. I changed a lot of things between the time I brought it down for rebuild, but it's about 15 mph faster now than it was then! So far, I've deployed it at about 130 mph, so it's clearly stout enough for the job, not that I was concerned about that...


Return to Mark Langford's N891JF Refurb Project