Purpose

This is a blog containing the build history of an experimental home built airplane. The RV-7A is a two place, piston powered, low wing, tractor configuration, tricycle gear, aluminum and composite aircraft. The original purpose of this blog was to document the construction of my experimental category aircraft in order to satisfy the build log requirement for the FAA. Now it's just for the amusement of friends and family as I document some of our aviation experiences. For more information on the RV series of aircraft see www.vansaircraft.com.

Wednesday, May 18, 2016

Step 14.0, Engine first start

Holy smokes that thing is loud!  That was the second thing that went through my mind when I saw the propeller turning on its own for the first time.  The first thing was just the surprise that it happened so quickly.  In retrospect, I guess there is no real reason to be surprised, but it seems that few things ever come along this easily and this was a glittering counter-example to everyday experience.

The one remaining task to perform before starting the engine was to check the fuel flow rate.  I wanted to be confident that when I finally turned the key, there would be fuel ready to ignite.  But in order to get fuel to the engine, I first had to have a place to store the fuel.  The fuel tanks are in the wings so I needed to put the wings back on temporarily.  I say temporarily because I can not transport the plane to the airport with the wings on. But on the other hand, I can't start the plane without the wings.  So on they go, for now.

I built a couple of saw horses that are at different heights to comp-ensate for the wing dihedral. This time around, and with the help of my new saw horses, I found the wing installation much easier.  After the wings were on I could move on to the fuel flow tests.  

To better understand what happens next, I'll explain the path that the fuel takes on its way to the engine cylinders.

The aviation fuel, 100LL, is held in the aforementioned wing tanks, each holding 21 US gallons.

The fuel pickup in each tank begins the fuel's journey to the engine.  The first stop is the fuel selector valve where the pilot chooses from which tank to feed the go-juice.  After the fuel selector the gas goes through the auxiliary electric fuel pump and through a maze of check valves, bypasses, and a filter.  It then goes on to the engine mounted mechanical fuel pump.  Either of these pumps is sufficient to power the engine, we have two for redundancy.  After the mechanical pump, the fuel travels to the fuel servo where it is metered using the mixture control and the manifold pressure. Following the fuel servo, the fuel then travels through the fuel flow transducer and then on to the flow divider, also called a spider.  The spider is so named because of its physical appearance is similar to its namesake.  The spider divides the fuel flow among the four cylinders.  The last stop on the fuel highway is the injector where the 100LL is atomized and inserted directly into each cylinder just ahead of the intake valve.

Center:  Fuel spider
The purpose of the fuel flow tests are to verify that the fuel pumps can supply enough fuel to the engine to sustain flight at high power and with a generous safety margin.  For my plane that number would be 125% of max flow at rated horsepower or about 21.35 gallons per hour.

Fuel test 1.
I broke the connection between the fuel servo and the spider and routed the fuel through a hose to a quart sized graduated container.  I then measured the amount of fuel delivered to the container over a period of 1 minute.  Simple, right?  Well, the fuel was just dribbling out and the fuel pump sounded like it was sucking air along with fuel.  This was troubling to say the least.  The amount of fuel collected after a minute was no where near that which would be required to sustain the motor under full power.

To trouble shoot the low flow rate , I started at the aux fuel pump because that is where I thought I could hear the sound of air being sucked into the lines.  There were no loose connections and no leaking fuel.  I then thought that perhaps the fuel pickup in the tank was sucking air.  But before I could test that I decided to move my attention to the output side of the aux fuel pump and check the flow there.

Fuel test 2.
The flow coming out of the aux fuel pump was quite robust.  In excess of 50 gallons per hour.  So the problem wasn't in the tank, fuel selector, or aux fuel pump after all.

Fuel test 3.  
I moved the collection point to after the mechanical pump, but before the fuel servo.  The flow here was awesome as well.  So now I was back to where I started, the problem was in the fuel servo.

This is when I had to change my operational method, and only as a last resort;  I would take a new tack, going forward, I decided to think a little before I tried stuff.  So it occurred to me that the fuel servo could not provide the flow I was hoping to see because its job is to provide fuel at the optimum ratio of air to fuel.  In my test there was no air component since the engine was not running and consequently, that meant that the servo was doing exactly what it should do -- provide only enough fuel to get the engine started.  Woo hoo! this thinking stuff was starting to pay dividends.

Well, a lot of hours were wasted chasing a phantom problem, but it's better than having a real one.  It was now time to actually start the motor.  After pulling the plane out of the shop, I removed the plugs and poured a small quantity of oil into each cylinder.



Then I rotated the prop by hand for a few minutes to try to get the engine's oil pump primed.  The plugs were then replaced and I positioned my pickup behind the plane to give me something to tie it to because I haven't been able to test the plane's brakes yet.

I have read many accounts of people having a great deal of trouble getting their engine going so I wasn't expecting much on my first try.   I even informed Carol that I would only let it turn for 20 seconds without starting before we would have to abort and let the starter motor cool down.   Carol's job was to video the event and check for leaks with fire extinguisher at the ready.




So as you can see, the engine started quite easily.  The white smoke at the beginning was just the oil that I had previously poured in to the cylinders.  There were no leaks observed and I was able the check the P-Mags.  In addition, the various sensors seemed to all be indicating correctly: oil pressure, fuel pressure, fuel flow rate, RPM, exhaust temperature, cylinder head temperature.  I cycled the prop control and verified the propeller governor was working.  After I looked at everything I could think of I shut it off.  Total run time was about 6 minutes.  So what kind of fuel flow did I actually measure?  I only let the RPM go to 2700 at which point I saw 16.9 GPH.  That is very close to the estimated value using HP*BSFC/6 = 17.03GPH.  More testing will be required and I will need to adjust the stops on the prop governor to limit the RPM.

Since I had the carpet out of the cockpit and since I had the canopy cracked open, my overall impression is that this thing is loud.  Crazy loud.  Like, dragster loud.  With four short pipes and no muffler it is not entirely unexpected, but jeez, this thing is scary powerful.  I was thankful that I thought to tie it to the truck because it was difficult to stop it from rolling. 


Finally, I'll wrap it up with another video that answers the question:
"Yeah it's an airplane, but what else is it good for?"




Answer:  Automatic tailgate lifter and bed liner extractor.

Monday, May 2, 2016

Step 13.9, Cowling

The cowling is the fiberglass enclosure around the engine.  It is analogous to the front quarter panels and hood of an automobile.  Theoretically, this cowling was made using molds taken from a perfectly fitting cowl from yesteryear.  If this cowl ever fit anything, it would be surprising indeed.  More on the fitment later.

The cowling is split laterally and each piece is supposed to be over sized such that careful trimming would result in one's own perfectly fitting cowl.  The first difficulty, though, is that no exact reference point exists in any dimension that would allow one to know where to trim.  However, I do know three things: first, that the back of the spinner must be about a 1/4" from the front face of the cowl. Second, that the opening for the prop shaft and the front is supposed to be 13" in diameter, and third, that the cowling is longitudinally symmetrical.

Since I didn't have the prop installed when I was fitting the cowl, my first datum, the back of the spinner, didn't exist.  No problem, I'll just mount the spinner back plate without the prop.  Now how thick is the prop hub.  Hmm, this could be trouble.


Here is the beginning of the fitment problems.  The upper and lower cowl halfs do not meet along the horizontal parting line.


No problemo, I'll just add more "side"to the lower front cowl.


Doesn't fit along the side either.  More cowl, please!


The upper front guides are glassed in.  These smooth the transition into the plenum.


A bit of spot putty fixes most of the unevenness.


I made these clips to hold the heater hose away from the exhaust pipes.


Now to cut the openings in the lower cowl for cowl flaps.  These openings allow additional air to flow, providing better cooling under high load or elevated ambient temperatures.




Now on to more sanding.  The reddish power is spot putty being sanded off.  The spot putty is primarily being used to fill the trillions of pin holes in the fiber glass cowl.


With the pin holes filled its time to apply the primer.



As it turns out, not all of the pin holes got filled.  I still have 150 billion to go.  I'm going to need more spot putty.

After one more round of primer, the cowl is finally ready for paint.


The next step is to install and test the cowl flaps.



Getting near the end now.  Adding the heat reflective foil to the inside of the cowl.  Hopefully, this will prevent the paint from bubbling on the outside.


Shush!  Sleeping puppy.


I'm not real happy with the final fitment of the cowling.  It seems that the fit is now different since the cowl has been painted.  Since that doesn't make a lot of sense I think I'll blame it on tolerance build up with the quarter turn fasteners.  Either way, the cowl doesn't fit as well now as it did before when it was just attached with clecos.  It's not obvious in the pictures, but there are some areas that will require additional work.


Also, the spinner is too close to the cowling so I will need to address that before first engine start. There are a lot of details to wrap up as I inch ever closer to finishing.  The old adage that 90% of the work occurs in the last 10% of the project applies here.

It seems that I am testing the Greek philosopher Xeno's paradox who noted some 2500 years ago that before one can finish his airplane he must first finish half of his airplane,  And before he can do that he must finish a quarter of his airplane -- and so on, in an infinite number of ever smaller steps that seem to prove that it is impossible to finish an airplane.  


Xeno was clearly ahead of his time!

Wednesday, February 3, 2016

Step 13.8, Spinner

The spinner is the little nose cone that covers the propeller hub.  It is not just ornamental, it plays an important part in smoothing the air flow over the cowling and guiding the air into the intakes on either side.  The construction is fairly simple:  A fiberglass party hat reinforced by a rear bulkhead and a forward bulkhead that sits about midway within the spinner.  How it got the name spinner shall remain forever a mystery :)

The rear bulkhead is delivered by Vans without the cutout.  Airplanes with fixed pitch props don't need the cutout, but for those of us with constant speed propellers -- well, we have more work to do. Cut a hole in the bulkhead to allow the propeller hub to protrude into the spinner cavity.   

                                   

Rivet on the bulkhead doubler.

Now test fit the spinner and bulkhead and drill out the mounting screw locations.  This takes a bit of measurement and calculation to get even spacing, but this step has gotten a lot easier since PI was discovered.


The next step is to cut out the openings for the propeller.


This is the front bulkhead bolted to the prop hub.


On a constant speed propeller, the blades twist to change their pitch during normal operation. Internally, hydraulic pressure moves the blades in one direction, while a big spring pushes the blades back.  Without the engine running, there is no hydraulic pressure so the spring pushes the blades back against their stops.  In the this step I will need to twist the prop blades to do the final fitting of the blade cutouts on the spinner, but the spring is really strong so I need some kind of leverage.  A couple of pieces of wood and two C clamps later I have a big lever.  The last part of the fitting is iterative: twist the blade until it contacts the spinner cutout, then make the cutout bigger. Repeat.



Here is where I run into the first problem with the propeller installation.  I get the prop mounted on the enginge and the prop bolts are lightly seated.  Great. Now I realize that these prop bolts need to be torqued to 65ft/lbs.  There's barely enough room for a thin wrench between the rear of the spinner and the flywheel/ring gear.  There is just no way that my torque wrench was going to fit in there.  There are probably expensive torque wrench extenders that I could buy, but I didn't find anything suitable after a quick look online.  This is where cheap tools are really valuable -- there is no need to feel bad about re-purposing them.  Carol found a $3 dollar combination wrench in some kind of bargain bin and I was able to cut off the end I needed and weld on a 3/8" socket adapter.  A new tool is born! Carefully measuring the extension allows an appropriate adjustment to be made to the final torque value using a ratio of total torque arm length to extension length.




So here we are with the propeller mounted and torqued.  Next, the prop bolts were safety wired.


It's a little bit hard to see what is going on here, but after the prop was mounted I wanted to see how the propeller tracked blade to blade.  A dial indicator is set up to measure the relative position of one blade and then the prop is turned 180 degrees to compare it to the other.  The two blades were virtually identical in their track, but one of the blades was .01" longer than the other.  I think I can live with that.



Almost finished now, the final step is to paint the spinner and then put it on your head.  Check.



All assembled, including a whole bunch of painted screws.







Friday, January 29, 2016

Step 13.7 Baffling


Given a choice, air would rather not cool your expensive airplane engine.  Due to its stubborn nature air must be confined and directed to flow past the cooling fins on the engine's cylinders.  This is not such an easy thing to do, but fortunately Vans has thoughtfully included a baffling kit within their 'Firewall Forward' kit. So what exactly is a baffle?  It is simply a box inside the engine compartment that has inlets in the front to let the cool air in, and outlets beside each cylinder to let the air out through the cooling fins.  The box must fit tightly around the top half of the engine so that all of the air passing through does its cooling work.  Much of the work in assembling the baffles is the trimming and test fitting process that tailors the baffle kit to a specific engine.  Let's get started:

In this photo I've just begun building the back corner.  The baffling kit is actually pretty good and the instructions are clear if not entirely complete and we shall see.


This is a hole I put in the left rear to connect a 4" hose to the oil cooler.


These two holes will be connected to become a spark plug wire pass through.



Here is a view of the right side mostly complete.


Time out to play.


A well know defect of the cylinder cooling design has fins that are too short or blocked near the base of the cylinders 3 and 4.  A solution shown below, it to pop out the box a bit to allow a little more air to flow by in this region.



One of the really annoying things about the ECI cylinder is that it has tapered fins that reduce in height toward the base of the cylinder.  This was probably done to reduce weight, but the baffle kit is not setup for these tapered fins.  Consequently, some additional pieces had to be fabricated to seal the box along the cylinder base.  As you can see it is a complex shape that takes a bit of time to discover by trial and error.  In this case, error and trial, seems more accurate.



Its just about impossible to get a good photo on these things installed.  The black item behind the baffle is the base of a cylinder.  Using some imagination, one can see how the complex shape forms a cap to fit the inside edge of the straight baffle to the bottom of the tapered cylinder.






It was about this time when I realized that the baffle instructions did not include the "snorkel."
The snorkel is the pipe that directs air from the front of the baffle box to the engine's air intake. Because the intake is below the engine this pipe runs up to the baffle it kind of resembles a "snorkel." Anyway, to fit the snorkel to the baffle, a big hole had to be cut out of the front left side of the baffle and then a flange was fabricated to attach the snorkel.  I really hated to take the snips to the nearly completed baffle, but I also really hate not finishing the plane...


I didn't take time to photograph the snorkel before I found out that it doesn't fit at all.  I had to make multiple adjustments to alleviate interference from the starter and from the alternator.  When I finished glassing in those changes I realized that it was just too long to mate up with the baffle box so I cut a half inch out of the middle and glassed it back together.  The photo below shows some of my handy work. The hole is a second stab at clearing an unused mounting tab that is inconveniently sticking out of the bottom of the starter motor.


After getting the snorkel to fit. I began work on a sliding door to act as an alternate air source.  The theory is that were the input to become clogged with ice or some other debris, it would be nice if the engine didn't shut off for lack of air.  The alternate air door can be opened in this unusual case to allow air in from inside the cowling.  Although it is unfiltered air, it's better than nothing.  The Vans supplied door has been found lacking -- the ability of close it once opened, for example,  So I thought I would try to design my own sliding door.  



And so here is the snorkel with the alternate air door completed.

A view of the snorkel from below after painting.


Air filter installed.


And finally the last step on the baffle project is to apply the fabric seal to the top of the baffle box. The seals press against the inside top of the cowling to prevent the air from escaping out of the top of the baffle box.