Author: William Wynne

Friends,

Here are a few quick notes on spark plugs. Print this off and keep it in your maintenance notes. I have a 3 ring binder that I keep in the top of my engine building tool box. In it I keep any data that I am not going to memorize. In my case, this is part numbers for things that we repeatedly order by phone, CC vs. compression ratio data, and research notes and test data.  Lots of stuff, like this plug data, I obviously have memorized, but the point is that well organized builders have notebooks and reference data, and it is a good habit to develop, especially if your workshop and home are not at the same place.

What plugs should I use? A common question. At our place, I often use Autolite 275s just to run engines on the ground. People have flown them, but the primary use I put them to is break in runs. They are in stock at most chain auto parts stores, and are often on sale for less than $1.50 each. I still like AC R44Fs for everyday flying. People have flown a giant variety of plugs, and the engine is not that sensitive to them with one exception: Do Not Fly an engine that will use 100LL fuel on platinum plugs. Other than this, make sure the plug you are looking at is the correct application. For many years the Bosch catalog listed for the Corvair a plug that was 3/16″ too long, and actually hit the piston head. If you are thinking of trying a different plug, go with one that people have already flown in a Corvair like yours, with the same carb and the same kind of fuel. For example, Woody Harris has a lot of flight time in his 2,700cc and later 2,850cc engines using Denso iridium plugs, part number IWF16-5359. His plane has an MA3-SPA carb and flies on 100LL. If you want to eliminate variables, use R44Fs, as they have proven to work well on the broadest variety of engine configurations.

How much torque do I put the plugs down to? This is a very important question. People used to cars with iron heads always overtorque Corvair plugs. The Corvair Shop Manual says you can use 20 pounds, and I have had new builders ask if 25 or 30 was ok, as they didn’t want them to “get loose.” If you routinely torque them that much on installation, they will get loose, because they are going to strip out of the heads. A much better number to work with is 7 to 10 pounds. I use 7 pounds more often than 10. After an initial ground run, I will recheck the torque.  The Corvair’s plugs seal by a gasket, and it takes almost no pressure on this to get it to seal. Don’t overdo it; it isn’t a lug nut on a diesel truck.

Above, final prep work on Lary Hatfield’s 3,000cc engine destined for service in his Zenith 750. I built the engine for him in our shop this week. It has all our Gold Systems and one of the Weseman’s Billet 5th bearings. After careful set up, the engine fired up after 3 seconds of cranking and laid down a flawless and smooth 1 hour break in run. Notice how short my personal plug wrench is. It is a 13/16″ plug socket with a hex top. I apply the torque with a cut down 12 point offset wrench that is only 4″ long. This arrangement fits in a small storage space. Because the wrench fits on up or down, it is very easy to use in confined spaces like the front two plugs without the Nosebowl removed. The bottle on the head is Champion plug lube.

What should I use for anti-seize?  There is only one substance you should ever put on any plug in an aircraft: Champion 2612. This is the only stuff that aircraft mechanics use. It is black graphite liquid with a tiny brush. It does a neat, controlled job. Over the years, I have seen a great number of planes of all types with plugs coated in silver anti-seize.  I have seen people apply it in the thickness one might better use to put peanut butter on a sandwich. Its brush is sized to apply it to diesel truck lug nuts, and the stuff is messy, and conductive. I have seen builders get enough on their fingers and on the ceramic part of the plug to cause a short, and make the plug boot slip off the plug. Stay away from it, get the real stuff. Aircraft Spruce sells small bottles that go so far that I am only on my second bottle in two decades of being an A&P. They cost less than $10.

What gap should I use? The ignition systems that I build are not too picky about it, but start with .035″. Measure it with a wire, and use an actual electrode tool to open the gap if required. Resist the temptation to pry the gap open with a screwdriver or a feeler gauge. As always, if you have any questions, give a call or write in.

Thank you,

William

Note to readers coming from the Cub Crafters site: This information is not directly applicable to Lycomings. The 6% power increase came from a major change in the intake configuration, not from the EFI. Look at 360 Lycomings; Yes most of the ones with a carb are 180HP and most of the injection are 200HP, a 10% gain, but this is ignoring the fact that the 200HP models (the -As and -Cs) have a very different angle valve head and a tuned induction system. Note that the 360s that just change from a carb to injection, (the -B series) remain at 180hp. There is no magic power in FI.  Also note that all Corvairs in the last decade all have electronic ignition. If a Lycoming has a perfect set of mags and they are replaced with electronic ignition, there is no reason to expect a measurable peak power increase.  The far easier way to get a small power increasein almost any motor without adding any complexity is to turn it slightly faster. -ww.

.

Friends,

Here are some notes and photos of five fuel injection systems for the Corvair. The first three are electronic systems, the latter two are mechanical injection. Fuel injection is a topic that many builders ask about.  I think a lot of the interest is generated by the awareness that high performance certified aircraft have mechanical fuel injection. A number of builders have learned that these systems are comparatively immune to carb ice. Beyond this, very few people have a good concept of the pros and cons of these systems vs. carbs. Even guys who know a fair amount about engines often miss important realities about the applications of these systems. This post gives a general overview, and covers details that are rarely discussed when builders bring up the topic of injection.

Electronic injection is the type of system that modern cars use. There have been a number of auto engine conversions that came from cars that had EFI that have gone on to fly in planes, and some suppliers to certified engines are just starting to look at EFIs fitted to Lycomings. What gives modern cars good mileage, long plug life, low emissions, etc.,  is the ability of the EFI system to operate in closed loop mode. It does this almost all the time the car is cruising down the road. When it is not doing  this, it is operating in “open loop mode” and falling back on the computer’s pre-programmed data that says ‘at 3150 rpm and 26.5″ MAP squirt in so much  fuel.’ In open loop, much of the advantage of EFI disappears. Up to here,  what I have typed falls into the category of  “Lots of people know this.”  Here is  the corollary only a few people understand.

In aircraft applications, the EFI systems almost never operate in closed loop. If you are going to cruise your aircraft at 75% power, it will spend its whole time in open loop. This is true with liquid cooled engines, but  really true with air cooled ones. Very few engines run with air fuel ratios of  14.7 to 1 at high power settings. They mostly run 12 to 1, or richer, and  O2 sensors have a hard time getting the loop to close at rich ratios. Sure,  there are exceptions to this, like wide band sensors, but you really cannot  compare a made at home system to a 2007 Corvette with perfectly tuned knock  sensors, 1 million lines of code in memory, and the ability to look at  individual cylinder exhaust pulses as they pass the O2 sensor. Even still,  GM knew that 75% sustained power in the Vette would be about 155mph, and the car  would spend .002 percent of its life there, so it’s OK if it is in open loop at that point.

After CC #9, I got Mark from Falcon to walk over to Jann Eggenfelner’s hangar. Jann is the king of Subarus, and like it or not, most people concede that he has flown more different types of EFI than anyone else. I know him  fairly well, and he is very smart, and unbelievably tenacious. With Mark’s OEM background and Jann’s flight experience, they had a very detailed high speed  data exchange. The recurring point that Jann kept coming back to is that no system, including the Subaru OEM stuff, will reliably operate in closed loop at aircraft power settings. In open loop, EFI begins to look like a very complex, high pressure, electrically dependent carburetor.

Above, the EFI 2,700cc Corvair built by Mark at FalconMachine.net in 2007, at power on my dyno. The urethane wheel directly reads foot pounds of torque off the digital scale. Note that this engine is using headers with collectors. We also tested it with cast iron manifolds and mufflers. It has distributorless ignition. Six LS1 coils are mounted on the sides of the black airbox. After a lot of careful calibration runs, this engine achieved a 6 percent power increase over a carbureted Corvair. Merely saying this will certainly activate the keyboards of armchair EFI experts, but it’s simple measured facts. Before questioning the test methodology or results, consider that Mark has earned his living with these systems for the past 20 years and the instrumentation included such niceties as a $500 laboratory grade digital oxygen sensor. Anyone who says that adding EFI to an engine like a Corvair will add 30% more power is just making their information up. The system above was tested a number of hours but was not flown. The controller on this was a Tracy Crook unit. This engine was equipped with equal length intake runners. It was laid out to fit in a 601/650 cowl.

 This is a redundant ignition, electronically fuel injected, fifth bearing, 2,700cc test engine built by Roy at RoysGarage.com in 2007.  It features coil on plug technology and throttle body injectors along with a rear mounted 40A alternator. It is mounted on Roy’s 701. It was run and tested, but not flown. Roy also has extensive experience with digital EFI systems. This provided good data, but in the end, Roy thought about the complexity he was applying to a very simple aircraft and chose to finish the aircraft with a simple gravity feed carb instead.

My thesis on EFI in 5 simple points:
1) Any system that uses lower pressure fuel is less likely to leak. Gravity is better than 5 psi, and 5 psi is better than 40. EFI runs at high pressure.

2) Any system that uses no electricity is better than one that uses  a little, and one that uses a little is better than one that uses a  lot, especially if the one that uses a lot needs it to be a  certain voltage, like digital EFI.

3) Any system that has less parts and connections is less likely to  fail. Digital electronic connections, working at low voltages, are very sensitive  to corrosion, temperature and vibration, things planes produce more than newer cars.

4) Almost all the things that EFI advocates hope for, HP increase, smoothness, fuel efficiency, and reliability, will prove elusive or minimal.  Before debating this, seek out a single flying system that will go into a closed loop in cruise flight. Realize that monitoring voltage and fuel pressure is not a work load reduction from using carb heat.

5) The only good reason to work on an EFI Corvair is because you want  a challenge, and this is more important than finishing your plane soon. This is  a valid position, and I support anyone who knowingly makes it.

Let me introduce a man who personifies point number 5, Rex Johnston. As far as I know, Rex is the first guy to ever fly an EFI system on a Corvair powered aircraft. This makes him someone special, because there were a lot of hurdles to jump over. His Corvair powered plane is a Davis DA-2, a sporty little two-seat sheet metal aircraft. Rex’s work is an outstanding example of building to meet a challenge that you personally feel. He was not after some illusive performance goal, he was just looking to challenge himself and learn a lot. Hats off to Rex.

Above, the underside of Rex’s plane. His system is a Holley Projection throttle body unit, that Holley originally sold for 258cid Jeeps. Rex’s  engine is a 3,100cc Corvair. Notice that it still has carb heat. A project like this isn’t for everyone. It takes significant experience building to be able to develop and flight test a complex set of systems like this.

The above photo shows an Airflow Performance mechanical fuel injector specifically calibrated for the Corvair. For size reference, a core Stromberg carb is at top left in the photo. Below it is the gold flow divider. One of the installation advantages of mechanical injection is the extremely small calibrated nozzles. Packaging six electronic injectors that will fit in a tight cowl is challenging. Mechanical injectors have an 1/8″ pipe thread on the bottom and are roughly 1/4 the size of electronic injectors. Airflow performance is owned by Don Rivera, a very smart guy who has owned and driven land-based Corvairs. This system is made in the U.S. The parts for this system cost over $3000.  Corvair builder Sarah Ashmore is putting one of these on her Personal Cruiser airframe. Her heads were modified for the injectors by Mark at Falcon. She works in the aerospace industry and made the choice to equip her aircraft with a system that met her specific goals.

Above is a photo of a Precision mechanical fuel injector.  The pictured unit is a port injector, but they also make a very compact unit that has the same quality, but has just one injector built into the body of the unit. The system does not require any modifications to the Corvair’s heads or intake, as the unit bolts onto the same flange as an aircraft carb. Peter Nielson of precisionairmotive.com is our engineering rep who has supplied us with a test unit, which we are now testing on a 3,000cc Corvair. This system is about $2,500. Precision knows quality, as they produce parts for and service certified fuel systems. We will release more data as we move through the tests. We are planning on flight testing this on Woody Harris’ 601, and Dan Weseman is planning on using the aerobatic capability of the system on the prototype of his new Corvair powered design, the Panther.

Today, 99.75% of Corvair powered planes use carbs. However, I think the Corvair is a good platform for fuel injection, and there are a small number of airframe applications that could really benefit from a fuel injection system. While there has always been a lot of talk about EFI, only a rare few clever and persistent builders like Rex Johnston will see the project through. I personally feel that the mechanical systems offer the best reliability and most proven track record, having flown in many demanding settings on other engines. As always, the proof and the progress is in the hands of the builders, the people In The Arena.

Thank you,

William

Friends,

The subject here is the most popular light aircraft carb of all time,the Stromberg NAS-3 and 3A. These were fitted to roughly 70,000 aircraft over a production run that lasted several decades. It looks like a very simple carb, and in some ways it is. But in reality, it is a highly engineered design that was produced by the world’s leading manufacturer of aircraft fuel systems.  At the time of production, this was Stromberg’s most basic product. They also made ultra sophisticated pressure carbs that were on the most powerful multi-row radials. This carb comes with an experience pedigree that no experimental carb can come close to matching.

Above, a Stromberg NAS-3 mounted on Dave the Bear’s Wagabond. We finished this aircraft in our Edgewater hangar in 2004, when Dave worked as part of  “The Hangar Gang.”  We used this combination with a 2,700cc Corvair and a Sensenich 64×35 prop to conduct a lot of jetting tests. We were not new to the Stromberg, as our Pietenpol, and our second test mule, Gary’s Skycoupe, both used the same carb. Besides these Corvair powered planes, the carb was also on several certified planes we had at the same time like Grace’s Taylorcraft and Gus’ 120.

When it comes to gravity feed carbs, I like Strombergs  because they have literally millions of hours feeding air and fuel into flight engines. I know them and trust them, and if I had any little issue with it, I’d have a mountain of expertise to draw on, not just other people flying one, but professionals with decades of documentation. I don’t have to think about it, it isn’t a variable. The one limitation on the carb is that it is not suited to use on a plane that requires a fuel pump. Thus, it isn’t good for a 601XL or a 650, but it is a good match for any high wing plane or one with a header tank. (The sole exception to this is the 750 because the factory now recommends that builders use a back-up fuel pump because of the plane’s high angle of attack capability.) I consider the Stromberg a much better carb for the Corvair than any of the other gravity feed carbs like a Monett Aerocarb, Posa, etc. The typical price for an old but functioning NAS-3 is $250. If you shop around, you can find them for half of this.

Above, an overhaulled NAS-3 that went on the Pietenpol of Dave Minsink.

The carb comes in two venturi sizes, but they are interchangeable, so if you have a core from a 65hp with a small 1.25″ venturi, you can replace the venturi with a larger 1.375″ one that came on the 85 and 90hp carbs, and you can rejet the carb to the bigger model. If you wish to have your core rebuilt, the best place we know of is D&G Fuel Systems in Niles, Mich. It is owned by a good guy named Russ Romey. I have worked with Russ over many years, and dozens of Corvair powered planes are flying with carbs that he rebuilt. If you don’t have a core, you can buy a Stromberg outright from Russ.

When Dave the Bear’s Wagabond was finished,  Gus did the original test flights. One of  the things we carefully worked out was the jetting. Here is something not to try at home: We did successive tests to slightly pull the mixture on full power climbouts, looking for a slight increase in rpm, which would indicate the carb running slightly rich to protect the engine from detonation. This was done carefully, and still Gus aborted a take off after reaching 150′ because the engine was hinting at detonation at the combination of timing and mixture  we were trying. Dave kept rejetting the carb until we got what we needed.  From the EGT reading, I think  that the correct A/F ratio at this setting was about 11:1.  

We spoke with Russ from D&G during this testing  and when we were done, he  began to jet the Strombergs he sells for Corvairs this same way. This is what we came to call a “Super Stromberg” to differentiate it from others jetted differently. The Corvair will run with one taken directly off a 65 or 85 Continental, but if you run it hard to get the engine’s full potential, it will be too lean. If you’re close to flying, and you know that you’re eventually going to have the carb redone internally, let me encourage you to do it before you start flying. Dave’s plane ran a whole lot better when we had it finished, and today people can just get their Stromberg to run this way right off the bat. The jetting works on any Corvair from 2,700-3,000cc.

Some models of the carb have a mixture control, others do not. The ones we used in our Pietenpol and the Skycoupe had no mixture. If I were planning on flying over 10,000′, I would have one, otherwise I would use either model.  Most of the light planes that originally came with NAS-3s only had a mixture control as an option, not standard equipment. Although this may sound a little crude for an aircraft carb, the mixture control isn’t a requirement for flying. People logged a lot of happy hours flying J-3 Cubs without a mixture control. 

One of the most common questions asked about Strombergs is their susceptibility to carb ice.  Although it is technically more prone to carb ice, I treat every carb as if it were prone to carb ice and I use carb heat on every plane and utilize it habitually any time the power is reduced below cruise setting. I consider the ice issue a small point compared to the other issues experienced by unregulated flat slide carbs. Sticking slides on planes that are on short final is far more of a risk . The Stromberg’s butterfly throttle plate always opens smoothly, and the carb is well known for holding an adjustment throughout a very wide range of temperatures and conditions. I have flown the NAS-3A on Grace’s Taylorcraft in temperatures ranging from 15F to 105F, and it does this without adjustment. Other than having the filter screen cleaned, the logs show that the carb on her plane has not been worked on since 1974. This is reliability.

Another point to consider is that the carb is very tolerant in changes to fuel density. A great number of certified planes that use these carbs have auto fuel STCs. This change is just a paperwork change that goes with the plane; there is no adjustment to the carb to start running auto fuel. Likewise, the jetting will not change when you alternate between auto fuel and 100LL in an experimental aircraft fed by a Stromberg.  This is not so with some experimental aircraft carbs. These two fuels have different densities, and some carb designs really need to be rejetted when changing back and forth between fuels. As much as I like working on planes, this is a lot of extra work compared to having a carb like a Stromberg.

The Marvel MA3-SPA is the only carb commonly used on a Corvair that has an accelerator pump. All the others, including the Stromberg, need a primer for starting. When the Stromberg has one, it will start in almost any weather you would choose to fly in. Grace’s Taylorcraft  engine is a C-85, but it is equipped with the exact same Stromberg carb we are speaking about. With no preheat, and the engine cold soaked at 20F, I can give it 3 shots with the manual primer (same type as in the Aircraft Spruce catalog #05-19920) , open the throttle slightly and crank the starter for no more than 2 seconds. The engine will light right off and run at a  high idle. ( I let it warm up for 5 minutes with the carb heat on. Not  only does this vaporize gas better, it does also slightly enrichen the  mixture by lowering the density of the air entering the carb.)  It was not  required to work the primer once the engine started, I just leave it locked. I  turn the carb heat off for a few seconds every minute to evaluate how well it  was working. Within one minute, turning it off gives a 75 rpm rise, indicating it is flowing hot air.  Even though they are different engines, this data should give people thinking of using the Stromberg a good indication that they will start easily when cold. On an actual Corvair, the higher compression and the stronger spark would make the engine light off even faster.

The Stromberg is an affordable, proven carb for gravity feed Corvair powered planes. I consider it an excellent carb for a Pietenpol, Kitfox, Highlander, Wagabond, and many other airframes. If you have a question about a specific airframe, drop me a note, I will be glad to share what I have learned.

Thank you.

William

From Jerry Tolman:

This blog is a great start to 2012. Thanks. I have a suggestion for a discussion…post a brief summary of the differences in displacement choices for those of us still early in the engine build or still deciding?
And a question…planning a West Coast Corvair College this year?

Jerry,

Good to hear from you. First, we are thinking of holding an event in California this year. We are looking at May 5th weekend in Chino at Steve Glover’s hangar where he runs NVaero. Steve is having a KR gathering there and an open house at his shop that weekend.  I have known Steve for many years, and his two commercial hangars at Chino are well equipped to have a very productive College.  Steve and I think that having the two events on the same weekend would be a good exchange of people and ideas. The Corvair is a very popular engine choice in the KR, and Steve wanted all Corvair builders to understand that they are more than welcome at his place. (Besides owning a series of KRs, Steve also has a Long EZ and a TriPacer; he is an all around aviation guy.) The only thing that has kept me from putting it on the schedule is a 50/50 chance that the weekend will conflict with my brother-in-law’s retirement from 30 years in the Army. In our family, such an event is “AHOD” (all hands on deck). We will know more in two weeks, and post the information here, under the “Events” heading.

On the subject of engine displacements: Today I encourage people to build one of three engine displacements that makes sense to them. Here is a breakdown:

 

2,700cc, 100hp, based on stock Corvair cylinders which are rebored from .020″ to .060″. Requires no machining to case or heads.  Uses either Sealed Power or Clark’s Forged pistons. This is the engine that 85% of builders are working on. More than enough power for a Pietenpol or a KR, a good choice for a 601 or 650. There are many flying examples of each of these airframes with this engine displacement that have individually logged hundreds of hours each.

2,850cc, 110hp, based on Clark’s heavy duty brand new full fin cylinders bored .105″. Requires no machining to the case or heads, it is a straight bolt-together engine. This engine uses U.S. made forged pistons that are available through us. We sell a Piston/Ring/Cylinder/Rod Kit for $1,750. The pistons have a 7.7cc dish in the head to lower the static compression while maintaining a very tight quench area, giving the combination of good combustion and outstanding detonation resistance, even on unleaded fuel. The engine is also the best choice for later turbo-charging. The displacement is a good choice for any Corvair powered airframe. Photos of CC #19 show Jeff Cochran’s running 2,850 built for his 750, and you can look at CC #21’s coverage for pictures of Clarence Dunkerley’s running 2,850 that he built for his Cleanex. Our Web site coverage of Oshkosh 2011 has a lot of photos of Woody Harris’ 601 that he flew on a circumnavigation of the U.S. Woody’s plane is powered by a 2,850.

3,000cc (3 liter) engine, 120hp. This is based on custom machined cylinder castings that are related to VW castings. The bore size is 92mm, but we use the HD casting that is the same as a 94mm VW cylinder. The piston in this engine is the big brother of the 2,850. It is forged in the same U.S. factory, and features a 10cc dish. This provides the same combustion characteristics in a slightly larger displacement.  This engine does require having the cases bored slightly for the larger cylinder spigots and having the head gasket area opened up slightly. This job must be done accurately, and the price is included in the Piston/Ring/Cylinder/Rod Kit. This displacement is a good choice for any Corvair powered aircraft. It has flown on both the 601 and 750.

Given these three engines, we no longer steer builders toward previous engines like the 3,100.  There is nothing wrong with 3,100s, but they did prove difficult to build for many people without a lot of previous experience. The internal dimensions were a compromise because the 3,100 used a modified VW piston with a different compression height, requiring the engine to be built with custom length pushrods, etc. The 3,000cc engine was our clean sheet of paper, based on what we knew after 10 years of building 3,100s. A 3,000cc engine is a better engine from a number of angles, and is a better engine choice for builders considering using unleaded fuel in the future. 

Thank You

William

Friends,

Every flying Corvair needs four pieces of cooling sheet metal that came from the factory. These pieces are the two under-cylinder baffles, and the two end plates on the firewall end of the engine that go between cylinder #1 and the distributor  and on the other side between cylinder #2 and the oil cooler. These four pieces of sheet metal are just as vital to the engine’s cooling as a radiator hose is to a liquid cooled engine.

Because a lot of engines have very filthy or rusty metal, we have decided to make sets available to builders. They are very time consuming to clean, even if you have a blasting cabinet. I took a large collection of these pieces to our powder coating guy, and let him professionally blast them, and then powder coat them matte black.

They are not required to be powdercoated on your flight engine, but getting a set from us does save some time, and they look very nice, a good compliment to a sharp looking engine. We are selling these sets for $99 plus your old parts by check or money order payable to William Wynne, 5000-18 US HWY 17 #247, Orange Park, FL 32003. If your core came without, please send us $120.

Please email us at WilliamTCA@aol.com to let us know they’re on the way. We’ll reserve a set for you.

As an introductory offer, we are going to cover the shipping of our parts to anywhere in the U.S. We will do this through the first 10 sets in the shop.

Above, the two end baffles in the bag, the two under-cylinder baffles are below them. All of the parts are powdercoated matte black.

Above, 14 sets of under-cylinder baffles ready to be sent as part of the four piece cooling sheet metal sets. In the background, freshly done powdercoated Valve Cover Sets.

Valve Cover Sets are available for shipping in the U.S. for $159 by sending a check payable to William Wynne with your old valve covers to 5000-18 US HWY 17 #247, Orange Park, FL 32003. Please email WilliamTCA@aol.com with your choice of Red, Blue or Black as well as your horsepower or FlyCorvair.com designation.

Thank you.

William

Friends,

When we assemble an engine, one of the steps that I take is to test the head studs before we put the case together.  It is a quality control step, and if one of the studs is slightly weak, I want to know it before assembly because it is a lot easier to fix before you build the engine. The procedure is fairly simple, and the tooling isn’t very elaborate. I bring my set to Colleges and show builders there the process in person. Here, in a few paragraphs and pictures, you can get a good overview.

Based on building several hundred engines in the past 20 or so years, the chances of getting a weak stud are low. About every  10th engine will have one. These studs were overstressed on disassembly or were overdone on a previous build. They look good on the outside, but the stud has been taken past its yield point.

Above, the test set up.The test is made easier with a high quality torque wrench, but it will work just the same with a beam type wrench. The small spacer on the arm allows the same tool to measure the longer top studs. The little tube is just a collar for the spacer, not required. The ends of the tubes need to be fairly true to the tube, the best method is turning them in a lathe, but careful work on a belt sander will do the same task.

Backing up a moment, I am going to assume that you have pulled and replaced all the studs that had hard tool marks from previous owners’ vice grips, and also pulled all the studs that have harsh rust pits. Mild surface corrosion is not an issue, and a missing thread at the top of the stud on the fine thread end isn’t a reason for rejection either.

Many years ago, I taught engineering labs at Embry Riddle in the Materials Department. People who have been through these classes know what a Tinius Olsen pull test machine is. For the rest of the gang, it is an immensely strong set of jaws pulled apart by a very powerful hydraulic system. Many of these systems can pull 50 thousand pounds without bogging down. A test sample of material is put between the jaws and very slowly pulled apart, while computers measure the length and power of the pull.  This all happens at a very slow rate, pulling 1/2″ can be slowed down to take several minutes.

Instead of demonstrating the machine on expensive test samples, I brought in bundles of Corvair head studs. We pulled them apart in every class. It gave both myself and the students a much better understanding of the effects of corrosion and mechanical damage like tool marks. A used Corvair stud in fair condition may not look that strong, but it takes over 10,000 pounds of pull to get one to neck down and break. Because of this testing I have a pretty good idea of what is too much damage on the outside of a stud. But the testing in these photos tells the condition of the stud on the inside of it.

The test tube is a piece of .188″ wall 4130 tube, 3/4″ in diameter. I welded a washer on the bottom it give it a bigger footprint. I polished the bottom of this so that it doesn’t leave any marks where the base gasket goes. On the top of the tube there is a very high quality hardened washer and an ARP 3/8-24 nut. (This can be done with lesser hardware, but remember, the goal is to test the stud, to the strength of the test hardware.)

Above, the washer and nut are in my hand, the tube is viewed end on showing the wall thickness. The tube has a little stand welded on it from a 2004 test series where we measured how much the studs stretched when they are torqued. At full torque, the studs are almost .035″ longer. This is an outstanding design feature. The engine is “spring loaded,” and the studs maintain their clamping force through a very wide range of engine temperatures, and expansion and contraction cycles. Engines like the Jabaru have very short bolts that hold the heads on. Bolts like that typically need continuous checking, because even a slight amount of material compression under the head of the bolt will result in a loss of torque on a short fastener. Conversely, a long stud is comparatively immune to this. Certified aircraft engines have the heads permanently screwed to their cylinders for a number of other design reasons, but engines like the Corvair, VW and Porsche all use the long studs. These are part of a well calibrated system, and are the primary reason why you should not use an aluminum cylinder on a Corvair. Porsche 911s eventually had aluminum cylinders, but they also had uber expensive “Delavar” studs, with an expansion and contraction rate that was compatible with their alloy cylinders. Companies that have offered aluminum cylinders for Corvairs have not taken expansion into consideration. Making the studs thicker or stronger actually only exacerbates the issue. Corvairs are designed for steel or iron cylinders, and they have an outstanding record of reliability with them.

Coat the threads on the stud and the washer with ARP Ulta torque lubricant. Drop the tube over the stud, run the washer down and then the nut. Carefully torque the nut to 15 foot pounds. Noting the clock position of the wrench handle when you start, raise the torque to 20 pounds. Typically this will require turning the wrench about 45 degrees on the short studs, about 55 degrees on the long ones. Next, raise the torque to 25 pounds slowly. Now, the critical observation: It should take the same 45 or 55 degrees of rotation on the nut to get the new torque increment. There is an acceptable range, and you shouldn’t be too concerned about a variation of 15 degrees or so. But, if you have a stud that requires 100 degrees of rotation to go from 20 to 25 pounds when all of the others took only 45 degrees, you have found a weak stud, and it needs to be replaced.

Above, the tool on a lower stud, giving a better view of the auxiliary arm on the tube from a previous test. The arm plays no role in this stud check up. Note the plywood under the case.  Don’t let the mating surfaces sit on a steel table, concrete or any other rough or hard surface.

A weak stud undetected is not going to lead to an engine failure. Typically, when a builder has a bad stud, he is torquing up his heads and notices that one stud turns way too far. This is the point where it would have been better to test before assembly. But even if it goes undetected at this point, the typical stud will not break, it just will not be clamping as tight as the others holding down the cylinder. In time this can lead to a blown head gasket.

Above, the task in action. The torque wrench is a $300 Snap On item, pricey, but an outstanding piece of quality. Ours is called Excalibur. If you ever meet an A&P mechanic and he has a pair of sunglasses or shoes that cost more than his torque wrench,  be guarded about taking his advice. Paul Gauguin’s paintbox was more valuable than anything else he owned. The brush doesn’t make the artist, and the tool doesn’t make the mechanic, but it is a measure of whether a man considers his work a craft or just a job. 

A Corvair is a very tough engine, and I have seen several of them fly a long way on a blown head gasket. The engine makes power on the cylinder even if the gasket is blown because at RPM the compression doesn’t have time to bleed through a tiny gap. A blown head gasket on a liquid cooled engine is a different story because it can mean a loss of coolant either out of the engine, into the crankcase, or into the combustion chamber. Liquid cooling is better in theory, but air cooling is better in practice. (A liquid cooled engine is less likely to ever blow a gasket, but if the discussion is about aircraft, you are mostly concerned about how the powerplant behaves after the event, not just the likelihood of the event.) In Corvairs, I have seen about 10 engines with blown head gaskets in the past 15 years. Almost all of these were caused by the timing not being set with a light at the static rpm. Only one or two were caused by a weak stud. Both of these causes are easy to avoid. Testing your studs before assembly avoids a small chance of a hassle on final assembly.

On final assembly, be alert for studs that take a lot more rotation to reach the rated torque value. When studs are torqued with ARP ultra-torque, we have done very careful tests to prove that you are getting the same clamping value at 26 foot pounds as a builder with light oil is getting if he torques the stud to 35 pounds. Use ARP, and stop when you get to 25-26 pounds. If you go all the way to 35 you are exerting a lot more force than required, and will actually be doing damage. Remember, you got into experimental aviation for the learning and adventure.  Take pride that successful Corvair engine builders know a lot more about how engines are really built than any other group in experimental aviation.

Thank you.

William