History of friction measurements at Airports

Airport Surface Friction Tester (ASFT)

The airport manager in many cases checked the friction conditions by making a skid test. If he was skidding too long he issued a ban on use of the airport. Due to accidents and incidents it was gradually found that better methods had to be developed for measuring friction of runway surfaces.

Why measure friction?

Flight Safety is the main reason for measuring friction. As the transport aeroplanes became larger it became also more important to check friction in a better way than making skid tests as mentioned above. Scandinavia, particularly Sweden, has taken a considerable part in the development of friction measuring technique.

Among reasons for friction measurements are:

  • Determine friction characteristics of runways under winter conditions

  • Verify friction characteristics of new or resurfaced runways

  • Assess periodically the slipperiness of paved runways when wet

  • Assess the effect on friction when drainage characteristics are poor

  • Assess friction of runways becoming slippery under unusual conditions

Development of the âKollerud methodâ for friction measurements

Scandinavian Airlines System, SAS, started 1946 to operateDouglas DC-4 aeroplanes on the then opened route from Scandinavia to New York. When in traffic the aeroplanes landed at the large military airportGardemoen.

For maintenance reasons SAS had occasionally to operate the large DC-4 aeroplanes also at the OsloFornebu airport. The main runway at this airport was then only 1200 m long with steep lopes at both ends. In order to avoid accidents at his airport the Airport Manager, Ottar Kollerud, started measurements of friction on the runway surface under winter conditions before DC-4 operation should take place.

Mr. Kollerud developed a method for friction measurements. According to this method a big truck was loaded with sand, accelerated to 30 km/h and then full brakes were applied resulting in locked wheels. Time and/or distance to a full stop were recorded.

From the recorded time, T, and distance,S, retardation, r, can be calculated:

r = V/Tr = V2/2S(r: m/s², V: m/s at brake application, T: s, S: m)

Kollerud reported the retardation in m/s². By test flights made by SAS it was found that the reportedretardation determined according to the formulas above corresponded to approximately half the value for the retardation of the aeroplane. The test flights were made with aeroplanes of the type DC-4. Later tests and calculations have shown that this is valid also for a lot of other aeroplane types.

The Kollerud method for retardation measurements is included in the ICAO Airport Services Manual, ICAO Doc 9137-AN/898, Appendix 5. The method is somewhat modified. In the ICAO document one instead of retardation calculates the friction coefficient,µ (mu). This is made by dividing retardation with g = 9.81 m/s². Simplified corrections are given to convert the measured friction values from µ (mu) skid toµ (mu) max.

Naturally, units can be in the ft system. Retardation will have a different value and g has to be in ft/s². US Air Force has used for friction measurements the James Brake Decelerometer,JBD, and that uses the ft-system. The JBD is corresponding to the Tapley- meter mentioned in Section 3.1.3 below.

It should be noted that according to the Kollerud methodskidding friction is recorded. When ICAOdiscussed friction coefficient measurements it was concluded that the maximum friction should be reported. This friction is recorded at a certain slip. Comments on this will be given later.

Need for friction measurements at other airports

The airport manager, Bertil Florman, at Bromma airport soon realised that he had need for friction measurements also at his airport under Winter conditions. At Bromma there are roads with intense traffic at both ends of the main runway.SAS and other Swedish operators found a need for friction measurements also at additional Swedish airports. SASwanted friction measurements also at Danish and Norwegian airports used by SAS. At this early time, late forties and the very early fifties, friction measurements had not been recognised as a problem internationally.

Early development of friction measurement technique

Mr. Florman started at Bromma airport operational friction measurements using the Kollerud-method. Soon it was found that that this method could be used because of the low frequency of DC-4 operation atFornebu, but that frequent use of the Kollerud-method was too time consuming and was ruining the brakes and the tires of the trucks.

Therefore, Mr Florman introduced the Tapley-meter for operational friction measurements. The Tapley-meter is adecelerometer that easily can be installed in a car. The car is accelerated and at a selected speed the brakes are applied. When the wheels had become locked and skidding the recording of the Tapley-meter was read.

The Tapley-meter method was also far less time consuming and its introduction was a great step forward in friction measuring technique. Friction characteristics were normally recorded at nine points along three lines, namely along the centerline and five meters on each side of this line.

Although the introduction of the Tapley-meter method was a great step forward in friction measuring technique Mr. Florman asked his friend the Chief Engineer, Mr. Kullberg, at the Swedish Road Research Institute, if he could develop a unit that would provide a continuous record of the friction along the runway.

Mr. Kullberg proposed to Mr. Florman to introduce hisSkiddometer method to record runway friction atBromma airport. The skiddometer method would mean that the maximum-friction was being recorded instead of the skidding friction, as up to that time had been the case with the Kollerud and Tapley-meter methods.

With a research Skiddometer, BV-1, Mr. Kullberg had shown as early as 1939 that on good summer roads the maximum friction for automobile tires is recorded at about 17 % slip, i.e. the peripheral speed of the braked wheel is 17% lower than the peripheral speed of the free rolling wheels.

An advantage of the Skiddometer method is that 80 to 85 % of the braked energy can be fed back to other wheels as a propelling force. During normal braking heat generation is a problem, especially when using the Kollerud method for friction measuring but also when using the Tapley-meter method.

Mr. Florman decided in the early fifties to introduce the Skiddometer for operational friction measurements at Bromma airport. A special Skiddometer, BV-2, was built. It was a trailer. SAS expressed the view that the skiddometer had to be a heavy vehicle in order to reasonably represent normal transport aeroplanes of that time. As a reasonable compromise was decided to load the measuring wheel with 1000 kg and the total weight of the SkiddometerBV-2 was 3000 kg.

The BV-2 had three wheels on the same axle. All three wheels had their own bearings and the shaft was equipped with two universal joints allowing the middle wheel, the measuring wheel, to have a smaller diameter than the two outer wheels. The diameters of the tires were chosen to result in a 17% slip.

As the braking force under good friction conditions can be500 to 600 kg or even more, if the load on the measuring wheel is 1000 kg, it was very important that 80 to 85 % of this force was fed back to the outer wheels and used to assist in towing the trailer.

The BV-2 was used at Bromma for many years for operational friction measurements.

Through the introduction of the Skiddometer method Swedish procedures were changed to measuring the maximum friction instead of the skidding friction that had been used up to the introduction of the Skiddometer method. Measuring and reporting the maximum friction is in line with ICAO procedures.

As the administrations of busy airports found that trailers had certain disadvantages SAAB started in the late sixties to develop a friction-measuring unit, the SAAB Friction Tester, SFT. A fifth wheel, the friction measuring wheel, was installed in the rear of a SAAB car model 99.

The measuring wheel was connected to the rear wheels of the car via chains and sprocket wheels. This means that the skiddometer principle is used and some 80 to 85 per cent of the braking force is used as propelling force. By selecting the teeth on the sprocket wheels and the diameter of the measuring wheel suitably the desired slip could be obtained. The slip of theSFT is 12%. This slip is selected for operational measurements in order to reduce tire wear.

The wheel load on the SFT is 140 kg. As in the case of the Skiddometer trailer further development of theSFT has taken place. The SFT is since long introduced in the ICAO, and FAAdocuments and ICAO changed the name SAAB Friction Tester to Surface Friction Tester, which also can be abbreviated SFT.

Calibration measuring

It was shown by tests made by the Aeronautical Research Institute, Stockholm, Sweden, that reliable calibration friction measurements results also were obtained, when the runway surface was contaminated by some loose snow or slush. Provided the friction measurements are made with a SFT with agrooved tire, with tire pressure 700 kPa and a test speed of 95 km/h is used. Now in Sweden, we have more than 25 years of experience, using this method. This technique is still the foundation for all ASFT CFME.

Early reporting technique

The early reporting technique was developed in co-operation between the Airport Authority at Bromma Airport andSAS. This took place in the early fifties. During a landing the friction characteristics of the middle portion and the far end of the runway are of primary importance. This led to reporting friction characteristics for three parts of the runway seen in the direction of landing. Soon the thirds were called A, B, and C.A is always called the low number runway end. An aeroplane landing from the high number direction got the report on friction in the order C, B, and A.

SAS and domestic Swedish operators understood what the friction numbers meant to them. However, operators coming in to e.g. Bromma airport did not understand what the reported numbers meant. Therefore, the expressions Good,Medium, and Poor were introduced.

SAS sent out a questionnaire asking for information from pilots on how they experienced information on braking action, i.e. friction, and also on controllability in crosswind.

About 3000 answers on these questionnaires were received. The answers showed that when a friction coefficient of 0.40 or above had been reported there were no pronounced problems on braking or controllability in crosswind. When 0.25 or lower had been reported the problems became severe. As a result of this study in Sweden was introduced the terminology:

Good 0.40 and above

Medium to Good0.36 to 0.39

Medium 0.30 to 0.35

Medium to Poor0.26 to 0.29

Poor 0.25 and below

As can be seen from the table we consider that no more than two significant figures should be reported. More than two figures would give a false impression of accuracy of the friction measuring equipment.

International recognition of Scandinavian procedures

An international recognition of the Scandinavianprocedures of measuring of friction characteristics at airports was when the Flight Safety Foundation awarded the Admiral Louis de Florez Flight Safety Award to the spokesman ofSAS and the Swedish Civil Aviation Administrationat numerous IATA and ICAO meetings since the early fifties. Saab friction testers have ever since helped make flying safer all over the world. Today all ASFTâs friction testers are using these principals and techniques.

page up