The problem is that the two rear cylinders run too cool and the forward two run too hot. After trying virtually every suggestion in the CP, and some others, with little success, Bill and Terry decided to do some serious testing and analysis of the problem. Using an airspeed indicator as a pressure gage (remember, an airspeed is simply an accurate pressure gage with the face marked in MPH or knots instead of PSI), six 1/8" ID clear plastic hoses were run from the cockpit aft through the firewall to various positions in the cowling. These hoses were numbered and tagged on each end and the cowling ends were reinforced with I" lengths of 1/8" OD brass tubing and securely lashed to various supports as available. The six locations tested were the top and bottom of the left two cylinders (4 places), just inside the NACA cooling inlet (5th place), and right on top of the per the plans installed oil cooler (6th place).
It really takes two people to conduct this flight test. Data was taken at a range of airspeeds and altitudes with OAT, CHT on each cylinder, oil temperature and engine RPM recorded for each set of pressure (MPH) readings. These data were then plotted up on graph paper as a function of altitude on one graph and airspeed on another. Careful examination of the numbers and graphs revealed that under all conditions tested, the rear cylinder, bottom side, consistently had the highest pressure while the rear cylinder, top side, had the lowest pressure.
Assuming all cylinders are externally essentially identical, with new identical baffling at the time of the test, then each cylinder has the same inherent resistance to air flowing through the fins. The pressure difference, bottom to top, across the forward cylinders, was much lower than the pressure difference across the rear cylinders. This results in much lower cooling air flow though the forward cylinders than the rear cylinders and, therefore, higher cylinder head temperature.
Almost all of the cooling air was going through the rear two cylinders. Basically, what happens is that the cooling air rushes in through the inlet, follows the bottom of the cowling as it swoops upward at the back till it hits the vertical rear baffle where this high velocity air is abruptly slowed down, raising its pressure. On the top side of the two rear cylinders, the lowest pressure exists due to proximity to the cowling outlet and the scavengering action of the prop. There is high pressure under the rear cylinders, low pressure on top and, presto, most of the cooling air flows through and around the rear two cylinders leaving the forward cylinders with less cooling air and much higher temperatures.
Obviously, the way to improve the cooling of the forward two cylinders was to increase the resistance to cooling air flow at the two rear cylinders. This was accomplished with some trial and error by installing temporary baffles forward of the vertical rear baffles under the two rear cylinders and cylinder heads to cover all but about 2" of the fin area of those two cylinders. With these temporary baffles wired in place, another flight test was conducted and instantly the CHT's were much closer to being even. One more iteration of even more restrictive, undercylinder baffles permanently solved the cylinder head problem.
The oil temperature problem, however, still existed on this 0-320 powered Long-EZ. Many ideas were tried. Some helped a little but nothing cured the problem until a second oil cooler was added on the right side. A "brute force" method to be sure, but one that worked incredibly well, although not too elegantly.
We would like to thank Bill Freeman and Terry Yake (both Long-EZ builders/flyers) for the above information and we can verify how well this method works on CHT problems based on personal experience. With a little "cut and try", all four cylinders can be within a couple of degrees of each other in level flight. Some differences still exist while in a steep climb but small compared to what we saw before. Obviously, it is essential to have a 4 cylinder CHT gauge installed in order to safely conduct these tests. Also, very important: keep in mind that, depending on the condition of the engine, indeed of each individual cylinder, you may have slightly different baffling requirements for your engine, or even each cylinder, than someone else has. Approach this test methodically and you will have excellent results.