Dicing With Icing: Smaller Aircraft Still Susceptible

Air Safety Week, Feb 28, 2005

With the number of icing-related accidents thus far this season, the Federal Aviation Administration's (FAA) planned safety meeting on Feb. 18 in Washington, D.C., to analyze the safety of corporate jets and on-demand charters must have been distracted somewhat by events nearly 1,700 miles away. Two days before the meeting, one of two Citation Vs traveling in consort (Reg. N500AT), had crashed in icing conditions during an approach at Pueblo, Colo., killing all eight onboard. The conditions were described as low cloud, freezing fog and drizzle.

A witness heard three distinct popping sounds just as the airplane went down at around 0913L. Initial speculation was that the airplane had flown through freezing drizzle and that the de-icing system may have been overwhelmed. In fact, typical of its class, this aircraft is not cleared for flight in severe icing. There have been a number of precedents.

It causes one to wonder just what precise icing mechanism is involved here. The Citation was the subject of at least two FAA directives on ice that required operators to modify planes or procedures. A March 1998 directive required a new warning to be included in the flight manual cautioning that freezing drizzle and other conditions could lead to an ice buildup that "may seriously degrade the performance and controllability of the airplane." An April 2000 directive required revisions to the flight manual and to a computer that calculates minimum safe airspeed.

According to a National Transportation Safety Board (NTSB) report on the Dec. 30, 1995, crash of a C560 in Eagle River, Wis., that jet was circling to land when it hit the ground about a quarter-mile from the runway threshold. The two pilots were killed; there were no passengers aboard. "The left wing and horizontal stabilizer leading edges had approximately one-eighth inch of rime ice adhering to their leading edges," the NTSB report said. Police reported precipitation in the form of freezing rain and sleet at the time of the accident, the safety agency added.

The NTSB said investigators found "both engines contained a small area of ice approximately 5 inches in diameter, which had formed beneath the final turbine wheel." The safety board attributed the probable cause of the accident to the failure of the pilot to maintain airspeed while executing the circling approach, along with factors that included "the descent below minimum descent altitude, the fog, the low ceiling and the icing conditions."

Giving Icing the Boot

On the Citation's wing and tail leading edges, cyclically inflatable deicer boots allow ice to build up and then shatter it. However, in common with many turboprops, the light bizjets cannot cope with super-cooled large droplets of freezing precipitation that accumulate all over the airframe. That type of icing just hits and sticks and builds up. The Citation's pneumatic boots cycle courtesy of 23 psi service bleed air. Besides OFF, it has 2 positions. AUTO does the tail then the wings via an auto-timer and MANUAL chucks hot bleed air to all four boots simultaneously (i.e., untimed). The inner wing panels in front of the engines are bleed air anti-iced, therefore, in theory, any "shed" ice won't be ingested by the engines if it's switched on in time (before the ice accumulates).

But there's the rub. Imagine that the system is in AUTO during the descent and initial approach. According to the NTSB and FAA's cautions, the system won't cope in severe rain-ice, so a layer will build up. When the pilots come visual below clouds and start their circling approach, they note the ice accumulation and switch to MANUAL to get rid of it. But at the same time they are "dirty" with gear and 15-- flaps and starting their level turn onto finals (for which they'll drop 35-- flap).

That's a lot of drag in a level turn, so they'd be simultaneously boosting the RPMs considerably because they know that the iced-up stalling speed in the base-turn is that much higher. The combination of changed angle-of-attack (due to flap and ice), higher IAS (indicated airspeed) and one other factor might have been enough to dislodge some ice from the inner wings. Why? Because those inner wing panels are bleed-air heated and suddenly at the higher RPM's the engines are belting out a higher volume of hotter air, possibly in continuous flow MANUAL (and the ambient temperatures are greater near the ground anyway).

Courtesy of the much reduced air-pressure over the inner wing panels in the flapped turn onto finals and the suddenly hotter air-heated panels beneath the thick layer of ice, the ice-sheet shatters and is sucked into the engines (which at the higher RPMs are sucking that much harder anyway). If N500AT was in a circling approach near Pueblo, Colo., that would explain the popping that was heard just before the aircraft suddenly dropped out of the sky at a rate, from the NTSB's initial findings (based on secondary radar height reports) of about 2,600 ft./min. over its final 30 seconds.

Here's another theory. Even without iced wings and tail, you can easily hit the pre-stall buffet due to urgently racking on the finals turn bank angle with a following (tightening) wind component -- in order to roll out on centerline. Any such stall buffet encounter would cause a pilot to instinctively go for high/max power and the shaking, wing-flexing and turbulent airflow over the wing would've helped liberate the inner wings' upper surface ice-sheets.