Friday, August 17, 2018
Transport Canada - Aviation Safety Letter "Damn, Was That Ever Slippery!"

"Damn, Was That Ever Slippery!"

by Paul Carson, Flight Technical Inspector, Certification and Operational Standards, Standards, Civil Aviation, Transport Canada

Runway excursion due to slippery runway

Runway excursion due to slippery runway

There isn’t a pilot in Canada operating high-performance aeroplanes who hasn’t uttered or heard someone utter something akin to the title of this article.

Operations on contaminated runways raise numerous questions from air operators. However, air operators aren’t the only ones operating under cold or inclement weather conditions interested in gaining a better understanding of the factors that influence aeroplane braking performance on non-bare and -dry runways. Their flight crews want to know more, too. While air operators are justifiably more concerned with minimizing the payload loss and maximizing their revenue, flight crews are more interested in maintaining a high level of safety in their own operation.

Hence, this article is directed at flight crews who want to know more about the why of contaminated runway operations than the what. The what, they are taught in the many ground school sessions they attend on the subject, generally at the expense of the why.

It doesn’t take a rocket scientist to figure out that a slippery runway affects the braking performance of an aeroplane. Any time you find yourself in a snowstorm, driving on one of Canada’s highways, you will automatically slow down, impelled by a strong sense of survival, because you instinctively know that it will take you a longer distance to stop. Guess what? The same is true for an aeroplane. In fact, even more so, because as all flight crews know, aeroplanes tend to make poor road vehicles! There are, of course, other issues that flight crews must consider when operating an aeroplane on a contaminated runway, such as loss in acceleration performance, if the contamination is deep enough on takeoff, say, or loss in aeroplane lateral controllability on a contaminated runway that just happens to be slippery and in a crosswind condition at the same time.

This article will not address all the aspects of operating an aeroplane on a contaminated runway. Instead, for the most part, the article will focus on the following: (1) what the measurement provided by a runway-friction measuring device means; (2) the difference between some of these devices; and (3) the difference between what these devices measure, that is, the difference between what is called a runway friction coefficient—or runway friction index or coefficient of runway friction—and the braking coefficient—or weight on wheels coefficient—experienced by an aeroplane. The runway friction coefficient and the braking coefficient are NOT the same thing, and this difference has lead to much confusion for flight crews because the manufacturers produce data using braking coefficient, while the airport operators report runway friction coefficient.

There shouldn’t be any conflict between operating an aeroplane safely and being economically viable in the process. In fact, it just makes good economic sense to operate safely at all times, while recognizing that to do so in adverse conditions may have an economic penalty. Pay it, and move on, or don’t operate!

Canadian Runway Friction Index (CRFI)— Application to Aircraft Performance

The information provided below, including information on the CRFI tables, (which have not been provided in this article due to space requirements) is drawn from the Aeronautical Information Manual (TC AIM) and can be found at the following Web site:

The data compiled in Table 1 (CRFI Recommended Landing Distances [No Discing/Reverse Thrust]) and Table 2 (CRFI Recommended Landing Distances [Discing/Reverse Thrust]) is considered to be the best available at this time because it is based upon extensive field-test performance data of aeroplane braking on winter-contaminated surfaces. Pilots will find the data useful when estimating aeroplane performance under adverse runway conditions. The aeroplane manufacturer is responsible for producing information and providing guidance or advice on the operation of aeroplanes on a wet and/or contaminated runway. The information below does not change, create any additional, authorize changes to, or permit deviations from other, regulatory requirements. The tables are intended to be used at the pilot’s discretion. Regulations and associated standards have been drafted on the use of the CRFI tables, and they are currently undergoing regulatory review.

Because of the many variables associated with computing accelerate-stop distances and balanced field lengths, it has not been possible to reduce the available data in such a way as to provide CRFI corrections applicable to all types of operations. Consequently, pending further study of the take-off problem, only corrections for landing distances and crosswinds are included.

It should be noted that in all cases the tables are based on corrections to flight manual dry-runway data and that the certification criteria does not allow consideration of the extra decelerating forces provided by reverse thrust or propeller reversing. On dry runways, thrust reversers provide only a small portion of the total decelerating forces when compared to wheel braking. However, as wheel braking becomes less effective, the portion of the stopping distance attributable to thrust reversing becomes greater. For this reason, if reversing is employed when a low CRFI is reported, a comparison of the actual stopping distance with that shown in Table 1 will make the estimates appear overly conservative. Nevertheless, there are circumstances, such as crosswind conditions, engine-out situations and reverser malfunctions, that may preclude the use of thrust reversing.

The landing distances recommended in Table 1 are intended to be used for aeroplanes with no discing and/or reverse-thrust capability and are based on statistical variations measured during actual flight tests.

Notwithstanding the above comments on the use of discing and/or reverse thrust, Table 2 may be used for aeroplanes with discing and/or reverse-thrust capability and is based on the recommended landing distances in Table 1, with additional calculations that give credit for discing and/or reverse thrust. In the calculation of distances in Table 2, the air distance from the screen height of 50 ft to touchdown and the delay distance from touchdown to the application of full braking remain unchanged from Table 1. The effects of discing and/or reverse thrust were used only to reduce the stopping distance from the application of full braking to a complete stop.

The recommended landing distances stated in Table 2 take into account the reduction in landing distances obtained with discing and/or reverse-thrust capability for a turboprop-powered aeroplane and with reverse thrust for a turbojet-powered aeroplane. Representative low values of discing and/or reverse-thrust effect have been assumed; therefore, the data may be conservative for properly executed landings by some aeroplanes with highly effective discing and/or thrust-reversing systems.

The crosswind limits for CRFI given in Table 3 contain a slightly different display range of runway-friction index values from those listed in Table 1 and Table 2. However, the CRFI values used for Table 3 are exactly the same as those used for Table 1 and Table 2 and are appropriate for the index value increments indicated. Further, it should be noted that the crosswind limits listed in Table 3 are not based on actual flight-test results, as are Table 1 and Table 2, because the hazards associated with such actual testing conditions were considered to be too great. To the best knowledge available, the results contained in Table 3 are based on a best estimate and have been available to flight crews in this very same format for many years.

Table 4 has also been updated based on the best data available, which was generated as a result of the testing program that helped produce Table 1 and Table 2.

Some additional comments about Table 1 and Table 2 are appropriate here.

Hidden in the tables is a middle step used in the development of the quoted distances. The first step was the correlation of the friction-measuring device used in Canada to measure runway friction, namely, a spot-measuring electronic decelerometer, to the μ (pronounced mu) braking coefficient of several aeroplanes that were tested during the winter runway contamination project. In order to develop landing distances in terms of the μ braking coefficient of any aeroplane, once certain values are assumed for the μ, all that is required is Newton’s Second Law—just some physics. The decelerating force is a function of the assumed aeroplane braking coefficient μ. Hence, once the correlation had been established between the measured runway-friction values and the braking coefficient of the tested aeroplane μ, the measured runway-friction values were used to calculate the stopping distances instead of some assumed aeroplane braking coefficient. Some manufacturers have stated positions indicating that it is not possible to take a runway-friction measuring device and correlate it well enough to the μ braking of an aeroplane. Based on extensive testing on winter surfaces in Canada and elsewhere, correlation coefficients over 90 percent have been consistently obtained for a wide range of aeroplane types. It’s time for minds to change!

The methodology used to derive CRFI tables is described in a number of reports published by the National Research Council of Canada (NRC). The methodology has also been adopted by a US-based standards organization: ASTM International. To a line driver, the preceding is just meaningless information. This information is provided only because there is a lot of technical literature available to those who want to dig for it. For example, during the production of CRFI, the researchers involved knew they were making a lot of errors—not mistakes of omission, but what we call known errors. These are errors that the researchers could do nothing about during the measurement process, but that had to be accounted for somehow. Using their best engineering judgment, the researchers decided to estimate what errors were being made and account for those errors in the final product you see as the CRFI tables. The generated data was heavily skewed to the lower friction numbers because that’s where the highest risks of operating aeroplanes on winter surfaces are found. When all the errors were added up, an accuracy level of close to 95 percent was achieved, which is why the tables come with the reported 95 percent level of confidence attached to them. In statistical analysis, this has a name. And for your next beer call, when you want to really impress, it is called a non-parametric statistical approach. Subsequent to this, statistical analysis was applied to the skewed data, what is called re-normalizing the so-called non-normal data to make it normal—nothing more than the familiar bell curve you used to get in university and college. It turned out that the data we had collected came in at over a 99 percent level of confidence. That’s the long way of saying that it’s pretty damn good data! Still, to account for the known errors being made, there is about 1 000 ft of error added in at the lower CRFI numbers and about 700 ft at the higher numbers to account for numerous factors, such as variation in friction cart readings across vehicles, friction levels changing on the runway, etc. This is all described in the early NRC reports referred to above.

How should the CRFI tables be used? This is a business decision that has to be made by every user. Linear interpolation within the tables is O.K., but it’s best to simply go to the next most conservative value. The tables are entered from the top with the CRFI value and from either the left- or right-hand columns with either landing distance or landing field length, as appropriate. For example, for a CRFI value of 0.32 and a dry landing distance of 2 500 ft, use 0.30 and 2 600 ft to avoid the interpolation. Extrapolation outside the tables is not recommended.

More needs to be said about the difference between landing distance and landing field length, the so-called 60 percent and 70 percent dispatch factors. There are many issues about aeroplane certification performance that today’s flight crews do not understand and that are simply not being addressed in any training material available to them or in a way that is understandable in terms of “pilotees.” One issue that is consistently misunderstood is the difference between landing distance and landing field length, which is described below.

There are operational dispatch factors that provide required landing field length and that are derived from landing distance. Note that dispatch factors or landing field length is an operational requirement, NOT a certification requirement, although some manufacturers include landing field length data or charts in the performance sections of their aircraft flight manuals (AFM), as noted earlier. Once the aeroplane is airborne, the dispatch factors no longer apply; only landing distance applies. For turbojet aeroplanes, the dry dispatch factor is calculated by multiplying the landing distance by 1/0.6 = 1.67. For turboprop aeroplanes, the dry dispatch factor is calculated by multiplying the landing distance by 1/0.7 = 1.43. As convoluted as the preceding appears, it is reproduced here because that is the way you will see it expressed in operational regulations. Most regulations on this subject, regardless of the authority that produces them, are almost incomprehensible. It is simply best to think of the numbers 1.67 and 1.43, as applicable, times the dry landing distance to obtain the dry landing field length. Clear!

How do you deal with an unserviceable component, for example, a zero flap landing? If it becomes necessary to apply corrections to the dry landing distance, simply enter the appropriate CRFI table with the normal, serviceable landing or field-length distance, as appropriate, to determine the recommended landing distance, assuming no unserviceable component existed. Then apply any additional corrections specified in the AFM for any aeroplane unserviceable component to the distance just obtained from the CRFI tables; otherwise, you will find yourself trying to use the CRFI tables outside the bounds. Again, extrapolation outside the tables is not recommended. The CRFI tables assume the anti-skid system is functioning normally.

Some concerns have been raised regarding the contaminated surfaces on which the testing was conducted to develop CRFI, the implication being that the results used to obtain CRFI are only applicable to those types of surfaces. The testing to obtain CRFI was conducted on mostly compacted snow and ice. These surfaces were used to obtain the desired low-friction numbers. Which other surfaces were we supposed to test on? CRFI is a non-dimensional number. It has no units and, hence, is not a function of the surface. If you get a decelerometer reading on some surface of, say, 0.2, then the CRFI tables would be applicable.

The presence of contaminants on a runway affects the performance of any aeroplane by (1) reducing the friction forces between the tire and the runway surface, (2) creating additional drag due to the contaminant impingement spray and displacement drag, and (3) leading to the potential for hydroplaning to occur.

There is a fairly clear distinction between the effect of soft contaminants and hard contaminants. The hard contaminants, like compacted snow and ice, reduce the friction forces only, while the soft contaminants, such as water, slush and loose snow, not only reduce the friction forces, but also have the potential to create additional drag and may lead to hydroplaning.

To develop a model of the reduced braking according to the type of contaminant is a difficult task, to be sure. That said, there is a runway-friction measuring device being used in Canada that has been successfully correlated to the braking coefficient of several aeroplanes, so that at least under certain contaminated runway conditions, such as compacted snow and ice, braking coefficient on a contaminated surface no longer has to be derived from some theoretical values on a dry runway—a highly suspect procedure at best. There appears to be no better substitute for actually measuring the value of friction on the runway and correlating that value to the braking coefficient of an aeroplane.


This article was published by Transport Canada in TP 185 Issue 4/2008 -. Reprinted with permission

Add comment

Security code