This report addresses glider towing accidents that occur shortly after takeoff when the glider climbs too high above the tow plane. A review of NTSB records from July 1983 to 1995 indicate only three accidents of this type, all fatal, occurred in the U.S. during that time period. Many similar incidents that happened at altitudes permitting recovery have been reported informally.

In most of the recoverable incidents the tow pilot reported the inability to release the tow line from the tow plane. In the three fatal accidents the tow line was not attached to either the tow plane or glider. A plausible explanation for this is that the tow pilot was attempting release but was unsuccessful until either the line broke or was released from the glider, thus relieving the stress on the tow hook. Release at this point was too late for recovery.

The type of tow hook commonly used for this purpose is the Schweizer 1D112 Series Tow Release Hook Assembly, installed in accordance with the requirements of Chapter 8, Section 2 of FAA Advisory Circular No. AC 43.13-2A. As shown in Schweizer Form F-236, the release effort should not exceed 60 lbs when a longitudinal load of 600 lbs is applied to the hook. However, in the incidents referred to above, the load probably was much greater than 600 lbs and applied at an angle to the longitudinal reference.

The study reported here was conducted in an attempt to quantify the factors involved in this kind of accident. It is hoped that a better understanding of what really happens may suggest ways to prevent them. Prior efforts along these lines seem to fall into two categories. One advocates better training of glider pilots and the other proposes some kind of automatic release for the tow plane. The virtual absence of automatic releases strongly suggests that all such proposals so far have been impractical. The fact that only three fatal accidents occurred in twelve years is testimony to the quality of pilot instruction as well as proof that it is not, and probably never can be, 100% effective.

Someone has said that one measurement is worth a thousand opinions. Measurements used in this study lack the precision one might desire, but it is believed they are accurate enough to permit drawing valid conclusions, subject to further refinement. The analysis utilizes information contained in several articles by John Roncz, published in "Sport Aviation" in 1990, as well as formulas from standard aeronautical texts. Review of the methodology is welcome. The three fatal accidents involved a Cessna 305A, Champion 7KCAB and a Piper Pawnee C. Data in this analysis is based on the Pawnee.

Figure 1 - Tow Hook Geometry

Since the tow line tension and the angle between it and the longitudinal axis of the tow plane vary significantly during normal towing operations, neither alone is appropriate for defining the load which jams the tow hook. Instead, the moment imposed on the pivoting arm of the hook ("A" in Figure 1) was used. It is directly proportional to the force between the pivoting arm "A" and the release lever "B" and a lot easier to measure and calculate. Direct measurements of tow hook moment, based on a sample of one tow vehicle and one tow pilot, indicate a value of about 1000 pound-inches or more jams the release.

Another attempt to simulate conditions in which the release jams is illustrated in Figure 2. It was still possible (although difficult) to release when the test was stopped at the values shown. No significant distortion of the release mechanism was observed during this test.

Figure 2 - Tow Hook Moment Test

When the tow line angle to the longitudinal axis is small the line will break before the jamming moment value can be reached. When the angle is large the maximum moment on the release arm is limited by the maximum net aerodynamic down force that can be produced by the tow plane tail. These conditions are shown in Figure 3 where the left part of the curve shows the maximum moment based on line strength. The right part of the curve shows the maximum moment limited by the aerodynamic down force on the tow plane tail.

Figure 3 - Tow Hook Moment Limits

Since the left part of the curve in Figure 3 depends on the tow line breaking strength, the problem of jammed releases can be eliminated by using a weak tow line. However, that simply trades one problem for another and might well cause more accidents than it prevents.

The right side of the curve, based on the aerodynamic force on the tail, is relatively independent of tow plane speed. The increase in dynamic pressure with increase in speed is offset by decreased downwash at the tail and increased negative pitching moments of the tow plane. The net aerodynamic down force at the tow hook is plotted versus tow plane speed in Figure 4. Note that as shown on both graphs, the net aerodynamic force at the hook is affected by movement of the center of gravity associated with loading. The two curves reflect loading extremes of a 125 pound pilot with zero usable fuel and a 200 pound pilot with full fuel.

Figure 4 - Maximum Tail Down Force Vs Speed

One approach to reducing the chance of the release jamming, suggested by Floyd Fagen, is to give the tow pilot the ability to exert a greater release force. If the release effort shown on Schweizer Form F-236 has a linear relationship to load, the release effort at the maximum moment shown in Figure 3 would be about 250 pounds. AC43.13-2A recommends a 5:1 mechanical advantage in the release mechanism, thus requiring the pilot to exert a 50 pound pull to release at the maximum jamming load. Actual installations of release mechanisms do not always provide a 5:1 advantage (typical Pawnee release levers provide slightly more than 1:1), but there are practical limits to how much the mechanical advantage can be increased.

When the Schweizer tow hook is used for auto tow, the "American Soaring Handbook" recommends mounting it on a hinge, "thus making it easier to free the cable in an emergency". This ensures that the tow line angle relative to the tow hook is always approximately zero and that the line will break before the jamming moment value is reached. It would appear feasible to adapt this type of mounting to aero tow.

Another approach to preventing jamming of the tow hook is to provide for automatic release when the moment reaches a predetermined level. Phillip Blum suggested replacing pin "C" in Figure 1 with one that would shear before the load reached jamming levels.

When used on tricycle geared airplanes the tow hook typically is mounted in the inverted position. In this case the moment on the release arm decreases as the tow line angle increases, suggesting that jamming could be prevented by simply inverting the tow hook. One incident involving a Cessna 182 tow plane supports this view.

Eliminating jamming of the tow plane release, while desirable, may not be a sufficient solution to the problem. In a typical incident the glider starts to climb and the tow pilot maintains pitch attitude by increasing up elevator. At some point with the stick fully aft the tail starts to rise. At this instant it still should be possible to release because the moment on the release arm is being relieved by the tow plane tail rising. However, the angle between the tow line and the tow plane longitudinal axis is decreasing - the tow hook moment is moving rapidly left and up along the right side of the curve in Figure 3. When the value reaches the line labeled "HOOK JAMS", it is no longer possible to release from the tow plane. Either the glider releases or the line breaks. In either case, or even if the tow pilot still could release, there may not be time and altitude enough for the tow plane to recover.

Automatic release of the tow line from the tow plane as soon as the tow plane tail starts to rise would be an good solution to this problem. William Hannahan has suggested an electrically operated system which would release when the tow pilot tried to move the elevator past its normal limits.

Similar results could be achieved with a mechanical device whose activation would be triggered by the horizontal tail reaching its maximum net aerodynamic force perpendicular to the longitudinal axis of the tow plane. It is the "MAXIMUM AERODYNAMIC DOWN FORCE AT HOOK" shown in Figure 4. Fortunately, it is relatively independent of speed. However, it does vary with tow plane loading. Assuming that the forces shown in Figure 4 are correct, a release designed to disengage at 700 pounds vertical force would prevent raising the tow plane tail. It also would allow the glider to climb to 30 degrees above the tow plane with 1400 pounds tension in the tow line before disengaging. This would seem to provide a reasonable margin of safety with respect to inadvertent release.

All of the solutions proposed in this report assume that the tow pilot is able to maintain pitch attitude by increasing up elevator until some limit is reached. In at least one incident, reported by Sonny Weller, the lifting force on the tow plane tail was applied so abruptly that the tow pilot had no opportunity to counteract it. This resulted in the tow plane being subjected to negative "g" forces and momentary total loss of control. In such cases, it is likely that none of the proposed triggering limits would be reached.

The complexity of many of the devices described here suggests that the most practical, if less desirable, approach would be one that at least ensures that the tow pilot can always release even when the tow line angle is large and the load high. A meeting involving six persons from the Arizona glider community and six persons from the FAA Arizona Flight Standards District Office was held at Scottsdale on November 9th, 1995. The group discussed all the proposals and eliminated most based on one or more considerations of reliability, probable cost or complexity of FAA approval.

The one which appeared to offer the greatest benefit for the least effort to develop and lowest cost to apply was selected for further study. It involved simply inverting the Schweizer tow hook so that the release load would decrease rather than increase as the tow line angle increased. Inverted releases are already approved for use with tricycle gear tow planes, and they apparently avoid the problem of not being able to release when the tow plane tail is lifted. Phil Blum designed an adapter to invert the tow hook, and installed it on his Piper Pawnee tow plane. As shown in the accompanying FAA Form 337, it consists of a short length of square tubing with the release hook mounted inside. The modified release performed as intended in a test at 1000 feet agl, and FAA field approval was promptly granted. Although this approach may not prevent all accidents of this type, it does at least give the tow pilot more control over his destiny when the glider climbs excessively after takeoff.

In addition to those named in the body of this report, the author would like to thank Timothy Borson, NTSB, Jack Christopherson, FAA, Roy Coulliette, Turf Soaring School, Gene Hammond, Soaring Safety Foundation and Bruce Stephens, Arizona Soaring for their contributions to this study.

Return to My Home Page