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 Collegesand 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.