Upper Wings #7

It's been nice to get back on airframe work!  Got lots of capstrip formed and installed.  The top sides of both wings are done.  

To recap (no pun intended) the process...

The curvature of the ribs is too much to attach the capstrips without pre-forming.  I soak the strips (3/16" X 5/8") in hot water in a PVC tube for about 4 hours.  It then goes into a form to dry.  This sets a curve which allows them to be installed without splitting and breaking. 

I also tried something different this time.  I pulled out all the nails on the lower wings front the capstrips after they dried (the nails are only there to hold the capstrip in position while the glue dries).  I wanted to make sure the nails didn't work their way out and eventually poke through the fabric.  Pulling the nails out was a pain, so I used little nailing strips this time to make the removal easier.  A lot of you are probably saying "DUH"!  Well, sometimes I'm just a little slow...

 Pictures below tell the rest of the story.

Capstrip material being soaked

Soaked strips in the forming jig

Strips installed with nailers holding them in place while the glue dries

Strips installed on the top side - our mighty Funk in the background
patiently waiting to be flown!

Detail of capstrips

Next will be repeating the same on the bottom of the wings, then it's trailing edge, leading edge, wing tip bows and finally the ailerons.

One other accomplishment these past couple weeks, I built up another Berling mag as a spare.  I was able to straighten the first armature I had re-wound that had about .010" runout.  Here it is...

Spare Berling mag ready to go

OX-5 Cylinders

Another chunk of work completed - the cylinders.  I took them to Mark's shop a couple of days ago to hone the barrels.  Mark made a nice honing tank that is set up with a Sunnen rigid hone.  This type of hone does the best job in keeping the bore round and even making it round if it is out.  

When I measured these prior to honing, I was amazed how consistent the barrel dimensions were.  The stock bore size is 4.000", and all cylinders measured within .001" of this anywhere they were measured.  Many, in fact, were within .0005".  I've had bored cylinders returned from a machine shop that were not this accurate and consistent.

So the honing was easy, all I had to do was take out a few minor imperfections - scratches, etc., and put in a nice cross-hatch.

Cylinder being honed with Marks very cool
Hudson project in background

Sunnen rigid hone

It's hard to tell in this picture, but the mounting base is not flat
due to being overtightened and not properly supported

I mentioned in the last post a little mod I made to the water outlet/rocker arm support on the top of the cylinders.

A poor design by Curtiss, along with many other things in this engine, but I just have to keep my head around the fact that this engine was designed over 100 years ago!  

The mounting base (shown at the bottom in this photo) is not supported at the ends where the studs go thru.  So as the nuts are tightened, the base bends.  As leaks occur and it is tightened down more, it leaks worse.  I've seen where some people flatten them back out by grinding or sanding, but I'm not a big fan of removing material that will actually weaken it even more.

So I came up with an idea (it's possible someone came up with this idea before me - forgive me if I'm not giving credit where it is due) to install an O-ring around the short tube protruding out the bottom, and chamfer the top of the cylinder to allow the O-ring to seal.  This has the advantage of eliminating the original gasket which will make for more rigid mounting since the rocker arm mounts here too.

The other problem to address are the support ears where the studs go thru so they don't bend down anymore.  I did this by machining thick washers that fit around the studs.  Now when bolted down, everything sits flat.  But first, I had to make sure the base was flat again. This was easier than expected by properly supporting it and using an arbor press.  

Here's the O-ring installed.  It's a Viton -018 size.  I chose Viton for its temperature range and chemical resistance properties.

I've used this modification before as a cylinder base gasket on engines that originally had no gasket.  Worked good in those applications, I see no reason it will not work here.







Here are the washers, the ones on top after being modified.  I bought extra thick, 1/4" washers even though the studs are 5/16". The O.D. of the 1/4" washer better matches the mounting base.  I had to machine them to about .092" thick and used a hole saw to notch them.  The I.D., of course, had to be enlarged to 5/16" too.

This picture shows the bore on the top of the cylinder after it was chamfered.  A chamfering depth of .095" gives the proper "crush" of the -018 O-ring.

It also shows the modified washers installed,  and you can see why the original design was prone to failure.  The area around the studs did not support the base of the rocker arm support, eventually causing it to warp and bend.  





It will be a while until I will actually know if this mod works, but I see no reason why it shouldn't.  I guess the one nice think about it - if it doesn't, I can keep the washers and install a regular gasket which will still be better than it originally was.

OK...for those of you bored with all this engine stuff, I'm bending capstrip and will soon be ready to start installing it on the wings.  Stay tuned for a wing update in the near future.

Berling Magneto Overhaul

Berling D81 X2

The Berling Magneto overhaul is done!  Before I get into the details, a little about the magneto.

The OX-5 used both the Berling and Dixie magnetos.  My engine came with the Berling, so that is what I went with.

The specific model is a D81 X2.  It's typical of early magnetos - horseshoe stationary magnets and a rotating coil/armature.  The rotating coil magnetos produced a good, hot spark, but were eventually replaced with the ones we know today that have rotating magnets and stationary coils.  It's my understanding that the rotating coils failed sooner due to the inertia and heat they were subjected to.

A magneto with 2 poles (horseshoe magnet like on the Berling) needs to operate at 2X engine speed on an 8-cylinder OX-5.  You can actually calculate the speed a magneto needs to operate at on an engine by using the formula:

Speed = (no. of cylinders / 2) divided by (no. of magnet poles)

We have to divide the number of cylinders by 2, since a 4-cycle engine makes 2 complete rotations per spark plug fire.  The distributor rotates at 1/2 magneto speed since we have to fire all 8 cylinders during the 720 degrees of rotation.

The Overhaul

Magneto in unrestored condition

The magneto I started with was original to my engine and was in unrestored, never overhauled condition.  So I was confident I had all the parts, I just wasn't sure what condition they were in.

Housing stripped and cleaned - ready for paint

The first steps in any overhaul/restoration is to clean and inspect the parts.  It's amazing when you look at these 100 year-old parts the effort that was made to make them.  There is no way in todays time that castings would be made that are this intricate - a beautifully made part!

Tool to remove the end cap of the adjustable bearing assembly

If you are going to take apart one of these magnetos, you'll need a couple of special tools that are used on the adjustable bearing assembly that supports the distributor gear.  They are easy to make.

The first one, shown to the right, removes and installs the cap that secures the 2 internal roller bearings.

Tool to remove the shaft nut

The next tool required is to remove/install the nut that holds the internal shaft in place.  The distributor gear attaches to this shaft, and the distributor finger goes thru the inside as well.  This is just a 1" socket that has the O.D. ground down to fit inside the housing of the bearing assembly.

Rewound coil/armature assembly

A couple of other components required special attention.  You may recall from previous posts that I had the coil rewound.  The original one checked out fine electrically, but I just don't feel comfortable flying in an aircraft with one magneto, that has a coil 100 years old!  Here you can see the rewound coil and the original laminated condenser replaced with a modern, solid-state type.  The spool to the left of the coil, is the collector spool that sends high voltage from the secondary coil to the distributor assembly.

Interrupter assembly

The interrupter (points assembly) was completely gone thru.  I disassembled and cleaned all the components and even had a new set of contacts to install.  You'll want to make sure the contacts are adjusted to make even and complete contact with each other. 

Cracked distributor finger locator bushing

One of the weak points of the 
Berling...the distributor finger and the locator bushing for the finger (shown to the left).  The finger is a long spindly shaft that needs to be inspected carefully for any signs of cracks.  The same for the locator bushing.  

It's hard to tell in this picture, but there is a crack that runs thru the mounting bushing and partially up the side.  If this were to come apart during operation - the magneto would fail.  Having spare parts, I replaced it with a good one.

 

The distributor finger (right), and locator bushing (lower left) are shown in this photo.  Inspect these carefully!

All the components, cleaned, inspected and ready for assembly

Finally, after all the cleaning, inspecting, replacing, etc. all the parts were ready for assembly.  The magneto goes together quickly and easily, but does require some care in a couple of areas which should increase the reliability of it.







Assembly

Shim being installed under bearing race

One of the first things to do is adjust the end-play of the armature.  There are no specs for this, but most magnetos I'm familiar with are adjusted with no perceptible end-play, but yet do not bind when rotated.  The end-play is adjusted with shims below the inner bearing races that are pressed on the armature shaft.  This requires a lot of fiddling around, as the shaft has to be installed, removed and re-shimmed multiple times to get it right.  Mine required a .004" shim to remove any end-play.

Internal timing marks

The magneto is timed internally, by aligning the marks on the armature shaft gear and distributor gear.

Clay being used to check clearance between distributor finger
and distributor

Distributor finger components showing nickel contact and shims

The Berling magneto utilizes jump spark distribution in the distributor block.  This means there is no direct connection between the distributor finger and distributor, the sparks jump - no different than in a spark plug.  Close clearance is important here so as much energy as possible goes to the spark plugs.  I checked this gap with clay, and adjusted the nickel contact in the distributor finger with shims.  I found the segments in the distributor block were not consistent, so I ended up with around .010" clearance, but no less.  I wanted to make sure there was never any contact between the two.

Puller required for removing bearing race

Oil groove machined into end plate to feed oil to bearing

The bearings in this magneto are manually oiled.  It's important the machined groove in the end caps are not plugged and feed this oil to the bearings.  The only way to know this is to remove the bearing races from the end caps.  You need a special puller to properly do this without damaging anything.  This picture shows a typical puller.  

When I got the bearing races out of mine, the grooves were packed with dried, thick oil residue.  These bearings would not have been properly lubricated.

Magnets being charged on magnet charger

With all these details out of the way, the magneto could be assembled and the magnets charged.  You can refer to a previous post on the magneto charger, but if you recall, I mentioned it was built to charge the magnets on the magneto.  That way you don't have to use a keeper on the magnets as they are installed.

My friend Mark bought a gauss meter (measures the strength of magnetic field) that he could use on his Model T magnets.  I decided to use it on this setup to see what effect the charging had.  Depending where on the magnets it was measured, the strength doubled from about 100 gauss to around 200 gauss after charging.  I don't really know if that is good or bad, but doubling seems good and the coming in speed would tell if that is adequate.


Completed Magneto

Completed magneto - ready to make the OX-5 Purrrrrrrrrrr!

Berling D81 X2 nameplate

Here are some shots of the final product.  She's a heavy one...weighing in at a whopping  19.5 pounds!  All that's left is to put in on the test stand and see if it will make some sparks.

On test stand with spark plugs

On test stand with electrode-type contact points

I tested the magneto with both the spark plugs I am going to use on this engine and the normal contact points that are on the test stand.  The magneto ran great, with a coming in speed of about 120 RPM.  I need to tweak the stand for more prolonged running as I would like to get several hours of run time on it prior to being used on the actual engine.


Final thoughts

If you read just about anywhere, the Berling magneto does not get much love.  Many people have opted for a more modern Scintilla or Bendix magneto, but those, being a 4-pole magnet, have to run at engine speed so a gear box is required to gear the magneto down.

I can say after overhauling this one, my level of confidence is greater than it was before.  Being able to carefully select and fit all the components will likely improve its reliability.  It is simple in construction, but yet has a lot of parts.  One thing I do like about it, is that it can easily be opened for inspection even on the engine.  So, we'll just have to see - time will tell!














Doing a little bit more cylinder work - honing the barrels and a little mod. to the water outlet that I will talk about on the next post.  Don't touch that dial!

OX-5 Oil Pump and Ignition Harness

Still waiting on capstrip material, so nothing is happening with the wings, so I keep plugging away on OX-5 stuff.

Thought I would take care of the oil pump.  Most people that work on piston engines are familiar with the term "blueprinting".  I guess you could say, that in some ways, this is what I did to this oil pump.

I have several different manuals that I refer to for OX-5 work:  The Curtiss manual, which is pretty basic, a manual from the Richfield Oil Company, which is much better, and one from Miller - a company that made OX-5 upgrade components.  It's interesting when you read what the Curtiss manual has to say about the oil pump.  "The oil pump should be opened and examined, although there is little chance of trouble here".  I have to take issue with this.  First of all, I would rate the oil pump as one of the more critical components of any engine - just like the heart is to the human body.  This is especially important with the OX-5 because of the screwy way the oil is sent thru the engine.  Oil is sent thru the camshaft first, then to the main bearings.  I feel it should be the other way around.  Secondly, I did find "trouble" with my pump!

Oil pump components

The pump is a simple gear-type pump made up of the components shown is this photo.  The first problem I found was the clearance between the bronze idler or driven gear and the shaft it rides on.  The clearance was .0079" - way too much!  The other shaft on the drive gear had a clearance of .0015" - acceptable.  So I turned a new shaft on my lathe with a .0015" clearance and pressed it in the housing.

Driven gear on new shaft with .0015" clearance

Checking clearance between gear teeth and pump cavity

Specs from the Richfield Oil Company manual call for .002" endplay of the gears, and .002" clearance between the gear teeth and pump cavity.  Either of these out of spec would affect the efficiency of the pump.

A piece of .001" shim stock revealed the clearance between the gear teeth and pump cavity very close to .002" - OK here.

Confirming end play with Plastigage

I wanted to dial in the end play as close to .002" as possible.  First was getting the gears to have the same height.  The bronze driven gear measured .3737" and the steel drive gear measured .3748", or .0011 higher.  I machined the drive gear to the .3737" dimension, so they now matched.  The cover was not flat, so I lapped it on a glass plate.  

After checking the depth of the pump cavity, it showed I needed a .008" gasket to get .002" endplay.  I made a gasket this thickness and checked the clearance with Plastigage and it came out right on the money!

May seem like a lot of work, but I thoroughly believe it is worth it!

Original harness

Then it was on to the ignition harness.  I need this done to test the magneto which I should be finishing soon.

I had the original harness, so it was just a matter of cleaning everything up and making new wires.  In my attempt to keep things as original as possible, I ordered 7mm, cloth-covered spark plug wire.

All the components ready for assembly

I had to make one new terminal plug housing to replace a
broken one

These are the original Rajah spark plug terminals.  Note the patent
date of Feb. 21, 1905.  Very Cool!!

Had to make a special tool to remove the plugs on the magneto
end

Installing terminals on the magneto end

Assembly is easy.  After the ends are stripped and tinned, it's just a matter of installing the terminals.  

On the magneto end, one half of the terminal is slid over the wire, a loop is formed in the wire, the housing slid over the end, and the plug screwed on with the special wrench.

Rajah terminal soldered on

On the spark plug end, the Rajah terminal is soldered on to the end of the wire.

Voila!  A finished harness

Not sure what's next.  I doubt I'll see the capstrip material, but my second coil/armature for the Berling mag is done so it may be magneto work!