Archive for the ‘Engine’ Category

Engine breather and oil separator

Sunday, March 2nd, 2014

Unless you drive some kind of classic antique, your car most likely has a system of plumbing and valves that takes crankcase blow-by gases and circulates them back into the engine intake to reduce pollution. On airplanes we just dump them overboard, because – as with many other aspects of aviation – that's how granddad did it back in the 1930's and by golly that's how we're going to keep doing it forever.

This takes the form of a "breather tube", which comes out of the top of the engine accessory case and is typically routed out the bottom of the cowl into the slipstream. As a bonus, the Lycoming engine also tends to spit a fine oil mist out the breather as well, which inevitably ends up all over the belly of the airplane.

At the source of the breather tube, my engine came with a straight 3/4" brass fitting that sticks out of the accessory case at an angle… and dead-ends right into the engine mount. Uh, that can't be good.

Another view looking aft. Not many options to attach a hose here.

The Van's solution to this problem is to use a pre-formed piece of rubber hose to take a sharp bend immediately after the fitting, just barely clearing the engine mount tube. This is the actual hose they include in the firewall forward kit, although you can obviously see the NAPA auto parts label on it. The problem is that automotive heater hose isn't compatible with petroleum-based fluids, which includes oil – it breaks down into a gummy mess after only a few hours. In fact, a Van's employee is even on record saying not to use this same material that Van's supplies. (Dear Van's, I love you guys, but you all need to get on the same page!)

I decided I would be better off using the proper type of hose, which necessitated a different fitting. First I had to remove the old one, which was very challenging given the limited space to work in. Actually it was tough to even get a wrench on it, and that sucker was in there tight. Finally after plenty of Liquid Wrench and the most force I could muster with rounding it off, the fitting came out.

And then after that, I fooled around for a week with various arrangements of fittings and tubing and hoses, trying to follow the plans method to get the breather line from the top of the engine to the firewall and down to the exhaust. The problem I kept running into was that the prop governor cable was always in the way, so I couldn't easily get the plumbing to go where it needed to without hitting something else. I think I probably could have come up with a solution eventually, but instead I started thinking, is there a better way to go about this?

Then, because I'm a glutton for punishment, I ordered an air-oil separator from the oddly-named Anti-Splat Aero company. The idea behind this system is that a centrifugal separator divides the air-oil mixture into its component parts, and returns the condensed oil vapor back into the engine. Out of respect for hallowed aviation tradition, the combustion gases still go overboard and pollute the environment, but at least the oil stays off the bottom of the airplane. At least that's the theory, we'll see how well it works. Field reports seem generally pretty positive, so I'm cautiously optimistic. The more immediate benefit to me is that this uses an entirely different arrangement of hoses, thus allowing me to take a different route around the problematic prop cable.

I don't like logo stickers, but fortunately theirs was easy to remove with acetone. Also, I ordered a white powder-coated oil separator, but they sent me the mirror-polished version. Whatever, it looks fine either way.

To protect the shiny finish I covered it with tape, which got progressively more dinged up as I experimented with mounting options and hose routing choices.

The oil separator needs to be mounted in a high spot on the firewall. It's pretty light, but I still didn't want to rely only on the 0.016" stainless firewall material to hold it up. I also wanted to use blind fasteners if possible, to ease future removal if required. So, I made up a little doubler out of scrap aluminum, and riveted a pair of nutplates to it.

To find the correct mounting location on the firewall, I first temporarily clecoed on the cowl fastener mounting strip. That let me figure out where I could position the separator, so it could be mounted as high as possible and still not interfere with any of the cowl fasteners that will eventually be riveted to this scalloped strip. Drilling the holes was yet another place where an angle drill came in handy.

I thought I might be able to use solid rivets to fasten the doubler to the firewall, but there's just too much stuff in the way to use a rivet gun and bucking bar. Instead I used Cherry Max blind rivets.

The oil separator clamps to an aluminum mounting bracket, which is bolted to the firewall via the nutplates that were just installed.

An important component of the whole system is this exhaust check valve and its associated mounting hardware. Instead of dumping straight overboard, the gas outlet tube from the oil separator is plumbed into the exhaust system through the check valve. This helps to burn up any remaining oil vapor before it can be deposited on the belly, and in theory also helps pull a mild vacuum on the crankcase due to interaction with the exhaust pulses. This may or may not have a beneficial effect on efficiency, but probably at least helps to keep oil leaks under control. The check valve is of course there to prevent the exhaust from pressurizing the crankcase.

The mounting tube faces aft and is cut at an angle relative to the exhaust stream, which helps pull gases into the exhaust pipe (see here for more information).

I had a suspicion that the valve supplied by Anti-Splat is the same as a NAPA part number 2-29000 smog valve, so I bought one just to take a look. It sure looks similar to me… supplied valve is on the left, NAPA valve is on the right:

They appear similarly internally also. This is good to know, since at eighteen bucks the NAPA valve will be cheap enough to replace every annual or two. Anti-Splat sells the (seemingly) identical part for $69, so if you were so inclined you could probably get by with ordering only the mounting tube and then buying yourself a valve at your local NAPA.

The innocent phrase "plumbed into the exhaust system", tossed out so casually above, implies a lot of head-scratching in order to try to find a place to mount the check valve where it doesn't run into anything else, followed by plenty of nail-biting and bad-word-uttering as you attempt to drill a bloody great hole in the side of a piece of expensive steel tubing without removing it from the airplane. I managed to get this done with a unibit and a long drill extension, but it was a pretty close shave.

Close-up of the hole. It's actually an oblong shape, due to the way the tube is angled. I used a dremel with a grinding stone to turn the round hole into an ellipse, followed by a scotchbrite drum to deburr and polish the edges. Also, apparently I left some of my blood behind. It's not the first time that's happened and I'm sure it won't be the last. Airplanes have sharp edges everywhere.

I didn't like the fact that there was no way to ensure the hole in the exhaust will be sealed, so I used a square of 1/16" thick asbestos gasket material between the mounting saddle and the exhaust. Well, not asbestos, but whatever they're using instead of asbestos these days. Unicorn hide, probably.

The check valve mounting pipe is attached to the exhaust with hose clamps. Hose clamps are safety-wired for security.

The gas discharge tube comes out of the bottom of the oil separator, which causes it to point towards the ever-troublesome prop governor cable. To turn it in a different direction, I used a plastic heater hose elbow from the local auto parts store. At first I was concerned that it might be too fragile for this application, but I had to manhandle it pretty roughly to get the hoses installed and it held up like a champ, so I'm no longer worried. The hose material here is MIL-6000, by the way… no radiator hose for me.

I also had to adapt the 5/8" hose outlet on the oil separator to the 3/4" hose required by the exhaust check valve… even though they are sold as a matching set. (why??!?) For this I used a brass 5/8-to-3/4 hose coupler just after the elbow. This whole business is a pretty Rube Goldberg way to turn a hose 90 degrees and step up to different size, but I couldn't find any single fitting that would allow me to do this in a way that was both more compact and also not any heavier.

The hose runs straight down out of the oil separator, then immediately turns 90 degrees and heads outboard, paralleling the prop governor cable. (The hose clamp shown here isn't tightened yet)

From there the hose is secured to an engine mount tube with a pair of adel clamps, and then it turns downward to the exhaust check valve.

Check valve and hose installed. There's a decent amount of clearance all around, and the hose doesn't rub on anything. Good enough.

Meanwhile, back to the top of the engine, where we still need to install a usable fitting for the breather outlet. I chose a 90-degree swivel elbow from a racing supply outfit. 1/2" NPT on one side, -10 AN flare on the other.

This fitting was pretty easy to install, since you can torque down the pipe thread portion without rotating the whole fitting. Without using a special swivel elbow, installation would be impossible due to the engine mount being in the way. Permatex #2 was used on the threads to avoid leaks.

The mating hose for the breather outlet is composed of a 90 degree hose end from the racing supply house – one of my many orders this month – and a length of Parker 801 hose. This stuff is a bit lighter and quite a bit more flexible than MIL-6000, while still being able to withstand exposure to oil.

The 90-degree elbow on the engine, combined with the 90-degree hose end, allow a hose routing that goes forward underneath the engine mount before turning aft again and heading to the oil separator. Another Rube Goldberg arrangement, but I couldn't find a better way to do it.

Another view from above. You can see here that the most obvious hose routing is blocked by the angle oil filter adapter, hence the somewhat convoluted approach I was forced to take. Further proof that the firewall forward area of the airplane is a zero-sum game… any item you install to make your life more convenient will cause exactly the same amount of inconvenience somewhere else! I suppose in theory it will all be worthwhile in the future when I'm changing the oil and the filter is so very convenient to get to. Someday. Maybe.

Two hoses down, one to go. The last hose required by the oil separator is a small drain line that returns the collected oil back to the engine. The recommended place to connect this hose is the turbocharger oil fitting on the side of the prop governor housing, seen here closed off by a brass plug:

It took a big socket extension, a flex adapter, and lots of elbow grease to get this plug out. I don't know what kind of sealant they used to use up at Mattituck, but it must have been good stuff.

The hole in the engine has 3/8" pipe threads, but the elbow fitting that goes into it is 1/8" NPT. Obviously an adapter is required, but I was unable to find one short enough to also allow me to install the elbow fitting without having it hit the engine mount. Argh! In desperation I sent an email to the Anti-Splat people asking if they had a source for a shorter adapter that would fit. I never got a reply, but a week later an envelope containing a low-profile pipe thread adapter unexpectedly showed up in my mailbox. Okay, I'll take it! Normal sized adapter on the left, special short adapter on the right:

Elbow fitting installed in the engine, following some careful work with a crow's foot wrench and a long extension, plus another dab of Permatex. If you have a fixed-pitch propeller you can probably clock it straight backwards, but on my engine you can't point it straight aft since the prop governor is in the way. After much experimentation I chose to angle it slightly upwards rather than attempting to route the hose below.

I initially wanted to use standard aircraft hose for the oil drain line, but the angled fitting required to work around the inconvenient prop governor would have made it expensive. Instead I made the drain line using more fittings from the race shop plus a length of fuel/emissions hose (SAE J30R7) from the local automotive parts emporium. If this doesn't hold up in service I'll use it as a template to order a proper aircraft hose.

The oil drain hose has a straight fitting on one end and a 45-degree fitting on the other:

The 45-degree fitting is necessary to allow the hose to angle out and around the prop governor:

The hose then goes up and around the P-mag, just missing… well, everything really. It's crowded back there.

The hose continues up to the oil separator, managing to miss everything else on the way. My friend John once described the firewall-forward section of the airplane as "a bowl of spaghetti where none of the noodles are allowed to touch", and I think that's about right. Too many hoses, cables, and wires running through a confined space, and none of them can ever touch each other or they'll saw themselves in half.

Finished! All hoses secured and clamped with Oetiker clamps:

…except this hose here, on which I used a normal band clamp due to space constraints:

What a relief it is to get this done and check another thing off the to-do list at last.

However, one thing to keep in mind is that I documented all of the above in a linear fashion, as if it all happened just that way. What I didn't show you was the weeks of uncertainty, frustration, false starts, backtracking, and scrapping of half-completed parts. This whole thing was a total ordeal, and pretty expensive to boot… since I didn't ever quite know what parts I was going to require, I ordered more than I thought I needed in the hopes that at least some of it would prove useful. Unfortunately I now have a big pile of leftover fittings which are going to have to go into my junk box miscellaneous parts container. Here's a representative sample of what was left on the cutting room floor:

Magneto P-lead / RPM sensor / oil temp probe

Sunday, December 29th, 2013

Three projects in one post today, for reasons that will become clear…

First, I fabricated the business end of the magneto P-lead. This connects to a switch in the cockpit that operates in reverse of the normal fashion. When the switch is open, the magneto can operate and generate its own sparks. When the switch is closed, the magneto is grounded and can't fire. For historical reasons unknown to me, the terminals on the magneto to which you connect this pair of wires are called "GND" and "P", hence "P-lead".

The open hole partially visible in the center of this image is where the oil temperature sensor goes. The hex-shaped thing below it is the exterior portion of the Vernatherm, which is a fancy trade name for "thermostat". The stuff at the top of this picture is where the oil filter threads on. Sorry for the terrible photo, this area is really close to the firewall so I can barely get my camera back there, let alone see it when I'm trying to work on it.

This is what the oil temp probe looks like. I crimped on a connector so I can replace it more easily in the future if I have to.

Oil temp probe installed with a new AN900 copper crush gasket. Crush gaskets are kind of interesting since you tighten them by turning to a specified angle rather than using a torque wrench. Since the hole uses a 5/8-18 thread, this gets tightened to 135° beyond finger-tight.

To install the RPM sender on the magneto, I had to remove the mag from the engine. Ugh! Wish I'd thought ahead a little more. Oh well.

The RPM sender threads into an unused vent hole on the mag, which is usually capped with a hex-shaped plug.

RPM sender installed in mag, using blue loctite and tightened to the specified torque. The P-lead and ground screw hole are visible at left.

Mag reinstalled… I used the timing pin and attempted to get it timed properly, although I will revisit the timing again for both the magneto and P-mag at some point in the future.

Now here is why I'm covering three otherwise unrelated wiring projects in the same post. All of these wires – P-lead, oil temp, and RPM – pass through the same area on the back of the engine, and all of them need to be supported so they don't vibrate and eventually break. The most convenient place to anchor these wires is actually the hole for the P-lead ground screw itself.

So what you're looking at here is the back of the magneto, with an adel clamp affixed to the hole that normally just gets a ground terminal; the ring terminal for the ground lead is between the screw head and the clamp. The other P-lead conductor loops around and attaches to the magneto "P" terminal, hidden under a rubber boot. Meanwhile, the wires for the oil temperature probe and the RPM sender pass through from left to right and are restrained by the clamp.

Again, sorry for the photo quality… this area is hard to see, hard to work on, and hard to photograph.

Moving upwards, the wire bundle loops around an oil hose and then attaches to the engine mount. Where there is currently masking tape holding things temporarily, I will eventually put a pair of adel clamps to grab onto this wire bundle. As with the bus current sensors, I used silicone tape around the connector for the oil temp sensor to keep water out.

Same area, this time with the oil filter installed:

Another glob of silicone tape to protect the connector for the RPM sensor, and then all three sets of wires pass through the firewall:

By the way, I did not think of this clever wiring trick myself – I got it from this VAF post.

Top spark plugs

Saturday, December 14th, 2013

Today I installed the top spark plugs on all four cylinders. For posterity, here's a photo of the box of plugs I used… NGK BR8EIX iridium plugs, as recommended in the P-Mag install manual. The 6747 part number has solid terminals, not the two-piece screw-on variety, which is preferred for this application.

The back of the box has a "not for airplanes" pictogram. That's encouraging.

The engine came with a set of cheap Denso copper plugs. I'm sure these would work fine, but I thought I'd go with iridium plugs based on past experience with an airplane that had a car engine in it. Also I've been really impressed with how long the OEM iridium plugs in my car have been going, with no appreciable wear.

View of the old and new plugs showing the difference between the two types of electrode.

You're not supposed to gap iridium plugs by hand, but I did check the plug gaps just to make sure they were manufactured correctly. The manual recommends a gap of between 0.030" and 0.035"… all four of my plugs were 0.028", which I'll consider close enough.

To use auto plugs in an aircraft cylinder, you need adapters to make the 14mm auto plugs for the 18mm aircraft plug bosses. When I sold my Lightspeed system, I kept the Lightspeed steel adapters rather than buying new ones from the P-mag people. That just means that I had to refer to the LSE documentation to install the plugs.

Using the LSE instructions, first the adapters get installed and torqued to 35-40 ft-lbs, using anti-seize and a new copper gasket:

Then the plugs are installed in the adapters, again with anti-seize, and torqued to 18-20 ft-lbs. This is different from the instructions for the P-mag adapters, which have you thread the plugs into the adapters and then torque them into the heads all as a single unit (to 18 ft-lbs). Anyway, since I have LSE adapters I decided to stick with the LSE instructions.

I put a light coat of silicone grease on the outside of the plugs to help the boots seat properly, and pressed the boots onto the plugs until they clicked:

I also re-did the ignition wire support arrangement a bit… I had previously piggybacked the support brackets onto the existing adel clamps that secure the fuel injection lines, but a VAF thread made me consider that such an arrangement could lead to increased vibration and potentially even cracking of the stainless lines. Probably a remote possibility, but just to make myself feel better I put in an extra set of adel clamps for the ignition wire supports, and restored the fuel injection line clamps to their previous configuration. Now there's no connection between the fuel lines and the plug wire supports.

Meanwhile, I haven't final-installed and torqued the bottom plugs yet, so used some red tape to remind myself they they are only temporarily installed:

Air duct bolts

Saturday, December 7th, 2013

With the alternate air door complete, it's time to bolt on the air duct for good. But first I had to take it off and remove the duct tape I put on there five years ago to keep stuff out… how time flies:

Photo for posterity showing what the business end of the fuel injection servo looks like – an inlet throat with a butterfly valve and a set of little air metering tubes:

The plans want you to use plain bolts plus lockwashers to keep them from backing out. I decided to use proper drilled-head bolts instead, just to be contrary. The part number for these is MS20074-04-06, in case you're wondering.

Safety wired the pair of bolts on the right side:

And here's the left side. Not my best safety-wiring job ever, but it's tricky to get the tools back in there.

The engine air duct is now permanently attached – bolts at the bottom and screws on top. That's a good feeling after all this time.

I installed the air filter to keep objects from falling into the engine, then taped a plastic bag over the top to keep schmutz from getting all over the air filter:

Alternate air cable improvement

Saturday, December 7th, 2013

I wasn't happy with the way the alternate air door could be opened with the cable, but not closed again the same way. After a couple days, though, I had an idea for a simple way to make it better. First I bought a package of spring washers from McMaster. These are just little pieces of round spring steel with a bend in the middle. You can get them in many different sizes and spring rates; these are the ones I used:

I made a little 1/4" OD brass bushing on my lathe, and enlarged the pivot hole in the door up to 9/32" so it would be a loose fit over the bushing. A quick trip to the hardware store for some brass washers and I was all set…

Here is the hardware stackup. Around a #8 AN525 screw is the brass bushing, surrounded by the two spring washers which are sandwiched between a pair of brass washers. A thin AN washer goes between the lower brass washer and the screw head.

Some blue loctite on the screw threads for extra security:

Here's what it looks like when it's torqued down. The screw captures the bushing between the outer washer and the steel door aperture; the bushing remains stationary while the door rotates around it. The outer washer compresses the springs, which in turn bear on the door to keep it from flopping outward when it's open. The brass washers are just there to provide a bearing surface for the steel spring washers as they rotate.

Here's what it looks like all assembled. In theory the moving washers will tend to try to loosen the screw, so this could still be improved by using a proper bolted joint with a castle nut and cotter pin. I already had the #8 nutplate installed on the inside of the air duct, though, so I'll keep an eye on this area and make sure it's lubed periodically. Between that and the loctite, I expect it should be fine.

The little thingy on the push-pull cable in the previous photo is one of these – a machined collar more typically found on model airplane landing gear:

I decided to put the collar on the wire to give it a positive stop. In retrospect I cut the cable sheath an inch too short, but the collar helps to make sure the cable doesn't pull too far out.

I thought about making a little video of the door being opened and closed with the cable, but it's too cold in the garage. Trust me, it now works in both directions! I'm happy with the way this turned out.