Saturday, October 18, 2008

Fuel Systems, AN fittings, NPT fittings, and their discontents

Follow this link, and scroll down a bit within it to read about a Zenair 601 with Corvair Power and it's fuel system at

I've purchased, read & digested William Wynne's books, both the Corvair conversion manual, and the relatively new ZenVair installation manual. He's made a pretty convincing case for the things he does by laying it down WHY he does things, and if you follow his site you can see the OTHER things he's tried.

Please examine the adjacent image, from the same web pace linked above. This is one of the cleanest fuel systems I've ever seen on an experimental, and I'd like to emulate it if I can on my airplane.

Let's follow this scheme from the firewall out. The AN bulkhead elbow (light blue anodized, AN833-6) is attached to a hand made "outie belly button" piece that spaces the deal far enough away from the firewall to hang the gascolator nice, and link up with the inlet to the primary Facet fuel pump. Later versions use a welded "firwall pass through", and you can buy this part from

A female-to-female straight union attaches the Andair gascolator which has male AN-6 connections built in. A 90 deg union connects the first Facet pump FP-40108, which also has in-built male AN-6 connectors. Another straight union, another pump, and on to the carb via a fuel pressure sensor.

Now, I've got two pumps, but they are the much more common Facet FP-40105. Can I devise something that works as well with the parts I have in hand, and can pick up from my local, well stocked hardware store?

The short answer is "NO".

The LONG answer starts with these facts:

1) The fittings I'd really like don't exist in NPT (National Pipe Thread). This means using more fitting/joints, which means more potential for leaks, especially since:
2) NPT RELIES on sealant, as it is NOT inherently leak free by design. No matter how tight you graunch down on it, there is always a spiral leak path around the threads.
3) NPT fittings SUCK from a fatigue standpoint, for several reasons. The mechanical function of tightening the joint, and the sealing function (such as it is, see answer 2) above) are not separated, and so if the joint fails mechanically in ANY way, it also leaks, not good with fuel. But the threads are also for shit, with sharp roots. True AN fittings have a special thread form and tighter tolerances than the nearly identical JIC 37 degree flare fittings, which are already way better than the NPT threads on hardware store fittings.
4) Even in the smaller 1/8 inch nominal diameter, the brass hardware store fittings are heavy compared to the aluminum AN-6 fittings - I haven't checked yet, but they may actually be heavier.

Nevertheless, I'm going to carry on with this plan, at least conceptually, for the following reasons:
1) A VW sucks less fuel than a Corvair. AN-6 is overkill at 4 GPM max.
2) I suspect I'll have to do a lot of experimentation to get my fuel system to fit, see the adjacent view of my firewall. I don't have a lot of real estate, and I may have to run the fuel pumps along the engine mount beds.
It will sure be a lot easier to mod the system if I can pedal my bike to the hardware store instead of going online to ASC or Wicks to get parts to play with.
3) The disadvantages of NPT fittings can be mitigated by NOT relying on them to support the gascolator, and by using Loctite Pipe Sealant.

So, my system, assume the same geometric layout, will require a 1/4" NPT bulkhead fitting, used in boat fuel systems, a 90 deg male-to-male elbow, a 1/4" female-to-female union, a 1/4" hex nipple, a Great Plain gascolator (the gascolator will be supported by a bracket, and NOT the NPT threads), another male-to-male elbow, another union, a reducing bushing, a 1/8" hex nipple, the first fuel pump, a 1/8" hex nipple, union, nipple, fuel pump, then off to the carb.

This fuel pump to carb transition bears examination. William Wynne uses an Earls AN-6 fitting with a 120 degree bend, connecting to HOSE, to deal with the vibration between engine and firewall. Nothing like this exists for 1/8" NPT, so I think I'll have to form something up using a NPT to 45 degree flare fitting, a suitable length ALUMINUM tubing with hose slipped over the ends of this and double clamped. Before you sneer, the HIGH pressure (70+ psi) fuel lines on our Volvo 164 used hose over straight pipe without a barb or a bead.

What I see as the VITAL points of William Wynne's system for a low wing aircraft, what I'm trying to retain: redundant fuel pumps, and no pressurized fuel in the cockpit.

UPDATE: The proceeding discussion can be viewed as an example of my thought process, or perhaps of the sort of mental masturbation first time builders with too much time on their hands, but not enough energy to go out to the shop, and, you know, BUILD anything, go through. The comment posters have convinced me AN fittings are the Way and the Light; the scales have fallen from my eye, and I BELIEVE. Amen.

Labels: ,

Saturday, May 03, 2008

Carb Cleaner is not for long term storage

About two weeks ago, I pulled my Tillotson cores out of the tub of toxic methyl chloride carb cleaner. They'd been sitting in there since, oh, last May.

Lots of parts were ruined. Both floats at the very least.

The pot metal pieces looked OK, but the brass parts seemed to have dissociated.

So, yes, let the stuff do it's work, but a month is probably as long as you should let carb parts sit in this witches brew.

UPDATE, October

I closely examined the pot metal upper and lower main castings, and decided they are junk. They look like they've been sandblasted. And the Carburator Refactory didn't have any Tillotsons for Model A's, only Bendix boat anchors.

So I'm giving up on the Tillotson X project. I don't want to fart around with them anymore, unless someone GIVES ME a NOS or completely rebuilt one before I can buy a Zenith 1821 from Great Plains.

IF you do find a nice core and want to try this carb, DO get the Master Rebuild kit, AND the book, and DON'T soak the pieces for longer than a couple of days before fishing them out, drilling out the plugs, reaming the passages, and finishing the rebuild using new gaskets.

ANOTHER UPDATE, later in October

I was cruising E-Bay and found a rebuilt, test run Tillotson X for $150 shipped. Bought it. Half the price of the new Zenith. And I have rebuild kits and a Grose Jet to spare!

Labels: ,

Tuesday, May 29, 2007

Notes on overhauling a Tillotson Model X Carburator

A Tillotson Model X Stripped Bare

Why bother farting around with a carb that's been out of production for at least 30 years? Well, some think that it's still an excellent choice for a flying VW engine. It's simple, robust, and does a pretty good job, for this size engine, of supplying the right fuel-air mixture over the entire speed range of use, which is about 800 to maybe 3600 RPM max, and that would be pushing it.

I bought a couple of cores, and recently set to work overhauling them. While there is a lot on the web about overhauling Zenith Model A carbs, there is bugger-all on the Tillotsons, which were/are MUCH lighter (Zeniths are CAST IRON), cheaper, and in some ways better.

First, a note about cores. It's best to examine them, or at least get good pictures, with the top cover off. One of my cores looks to be unsalvageable, as the boss for the main jet is broken, right in the throat where it all happens, mixture wise. I might be able to fix this with some JB weld, but for now, it's paperweight. And neither core I bought online had a choke/mixture driver. Not essential for my aviation application - I'll be using a primer to cold start the engine - but if you're building up a car, you'll want one. The OTHER core I bought had it's own problem, it looked like The Jolly Green Giant had stepped on the intake, more no that anon.

Another issue is the overhaul kits. Some are MUCH more complete than others. The $75 kits have damn near everything, including new choke and throttle shafts and butterflies. My $15 kit has the gaskets, jets, float valve, the several needle valves, and not much else. But it's sufficient, with my cores, although I would like to have the smaller passage plugs to be able to clean out EVERYTHING.

Finally, the 13/32 inch wrench you'll need to get the Gas Adjusting Needle Housing out is a rare beast indeed, get a cheap 10mm open end wrench and file it to fit.

Disassembly is pretty self explanatory. If something doesn't want to come apart, soak the whole deal in carb cleaner a day or so before trying it again. You can use a LITTLE heat on the housings, but not much, or you'll have a shiny puddle on the shop floor. Rust is not a problem with Tillotson's, but old dried up gas is pretty effective locking compound. So, back in the dip tank.

Here is the problem I had with my "best" core. What's wrong with these pictures:

Notice the circular crater in the center of the intake horn? (this is the top of the lower casting by the way) Some Brain Donor in the past honked down on the grub screw holding the Air Maze on so hard it ovalized the intake horn itself. It was a lot worse before I took these pictures, I've already pressed it almost back into shape.

Here you can see the bottom of that crater protruding into the ID of the "kirksite" casting, also called "pot metal", a low melting point mixture of zinc and aluminum that is a pretty great material for this sort of casting, as long as you keep it away from salt water. The choke shaft and butterfly plate goes in here, the main jet is way up in there, and the Gas Adjusting Needle Housing is still in this core, it's the brass tube on the right side. I'll use my fancy tooling to get this horn close(r) to round again.
Here's the fancy tooling, a 1 inch thick piece of (hardwood?) plywood scavenged out of an old couch. The idea is to press it back into shape, without putting too much of a point load on the rather brittle casting. Note the "tooling pins", other wise known as common nails with the heads cut off. I use a lot of them. Thanks, R.S Hoover.
Proceed slowly and cautiously please. I only have one good lower casting, and this is more of "Nice To Have" action, as the carb heat box I'm planning to make up soon will be able to accomodate the out-of-roundness that's left at this point.

Most homebuilt carb heat boxes use the nice spun aluminum flanged fittings you can get from Aircraft Spruce. But the Tillotson throat doesn't match any of the available sizes exactly, so I've been scratching my head, and consulting my reference books for a way to custom fabricate such a thing without welding or machine tools. Found it, look at this:
This flanged fitting is two pieces of sheet metal. The inner tube penetrates the wall of the box, presumably into a tight, flycut hole. The outer tube has the flange. It looks like there is a bit of spooj sealing it all up. I can make THAT. The rest of this guy's engine installation looks really good, check it out. And while you're at it, read a little about R.S. Hoover's approach to carb heat boxes, and William Wynne's personal experience on why you need one.

Labels: , ,