Appendix 2 -- Test specification rationale

• ED Decision 2018/007/R found in: CS-27 Amdt 10 - Small Rotorcraft (Feb 2023)

Appendix 2 — Test specification rationale ED Decision 2018/007/R (a) Introduction The overall risk of capsizing within the 5-minute exposure period consists of two components: the probability of capsizing in a given wave condition, and the probability of experiencing that wave condition in an emergency landing on water. If it is assumed that an emergency landing on water occurs at random and is not linked with weather conditions, the overall risk of a capsizing can be established by combining two pieces of information: (1) The wave climate scatter table, which shows the probability of meeting any particular combination of *Hs* and *Tz*. An example scatter table is shown below in **Figure 1 — Example of all-year wave scatter table**. Each cell of the table contains the probability of experiencing a wave condition with *Hs* and *Tz* in the range provided. Thus, the total of all cells in the table adds up to unity. (2) The probability of a capsizing in a 5-minute exposure for each of these height/period combinations. This probability of capsizing is different for each helicopter design and for each wave height/period combination, and is to be established through scale model testing using the method defined above. In theory, a model test for the rotorcraft design should be performed in the full range of wave height/period combinations covering all the cells in the scatter table. Clearly, wave height/period combinations with zero or very low probabilities of occurrence might be ignored. It might also be justifiably assumed that the probability of capsizing at very high wave heights is unity, and at very low wave heights, it is zero. However, there would still remain a very large number of intermediate wave height/period combinations that would need to be investigated in model tests, and it is considered that such a test programme would be too lengthy and costly to be practicable. The objective here is therefore to establish a justifiable method of estimating the overall 5-minute capsize probability using model test results for a single-wave condition. That is a single combination of *Hs* and *Tz*. Such a method can never be rigorously linked with the safety objective, but it is proposed that it may be regarded as a conservative approximation. (b) Test methodology The proposed test methodology is as follows: The rotorcraft designer selects a desired significant wave height limit *Hs(limit)* for ditching or the emergency flotation certification of his helicopter. Model tests are performed in the sea condition *Hs(limit) Tz(limit)* (where *Tz(limit)* is the zero-crossing period most likely to accompany *Hs(limit)*) with the selected spectrum shape using the method specified above, and the 5-minute probability of capsizing (*Pcapsize*) established in this sea condition. The way in which *Pcapsize* varies for other values of *Hs* and *Tz* is not known because it is not proposed to perform model tests in all the other possible combinations. Furthermore, there is no theoretical method to translate a probability of capsizing from one sea condition to another. However, it is known that the probability of capsizing is related to the exposure to breaking waves of sufficient height, and that this is in turn linked with wave steepness. Hence: (1) the probability of capsizing is likely to be higher for wave heights just less than Hs(limit) but with wave periods shorter than Tz(limit); and (2) the probability of capsizing will be lower for the larger population of wave conditions with wave heights less than Hs(limit) and with wave periods longer than Tz(limit). So, a reasonable and conservative assumption is that on average, the same Pcapsize holds good for all wave conditions with heights less than or equal to Hs(limit). A further conservative assumption is that Pcapsize is unity for all wave heights greater than Hs(limit). Using these assumptions, a comparison of the measured Pcapsize in Hs(limit) Tz(limit) against the target probability of capsizing (Pcapsize(target)) can be performed. In jurisdictions where flying is not permitted when the wave height is above Hs(limit), the rotorcraft will have passed the certification criteria provided that Pcapsize ≤ Pcapsize(target). In jurisdictions where flying over waves greater than *Hs(limit)* is permitted, the rotorcraft will have passed the certification criteria provided that: *Pcapsize**≤ Pcapsize(target)**– Pe*, where *Pe* is the probability of exceedance of *Hs(limit)*. Clearly, in this case, it can be seen that it would not be permissible for the rotorcraft designer to select an *Hs(limit)* which has a probability of exceedance greater than *Pcapsize(target)*.  Figure 1 — Example of all-year wave scatter table [Amdt No: 27/5] Appendix 2 — Test specification rationale ED Decision 2018/007/R (a) Introduction The overall risk of capsizing within the 5-minute exposure period consists of two components: the probability of capsizing in a given wave condition, and the probability of experiencing that wave condition in an emergency landing on water. If it is assumed that an emergency landing on water occurs at random and is not linked with weather conditions, the overall risk of a capsizing can be established by combining two pieces of information: (1) The wave climate scatter table, which shows the probability of meeting any particular combination of *Hs* and *Tz*. An example scatter table is shown below in **Figure 1 — Example of all-year wave scatter table**. Each cell of the table contains the probability of experiencing a wave condition with *Hs* and *Tz* in the range provided. Thus, the total of all cells in the table adds up to unity. (2) The probability of a capsizing in a 5-minute exposure for each of these height/period combinations. This probability of capsizing is different for each helicopter design and for each wave height/period combination, and is to be established through scale model testing using the method defined above. In theory, a model test for the rotorcraft design should be performed in the full range of wave height/period combinations covering all the cells in the scatter table. Clearly, wave height/period combinations with zero or very low probabilities of occurrence might be ignored. It might also be justifiably assumed that the probability of capsizing at very high wave heights is unity, and at very low wave heights, it is zero. However, there would still remain a very large number of intermediate wave height/period combinations that would need to be investigated in model tests, and it is considered that such a test programme would be too lengthy and costly to be practicable. The objective here is therefore to establish a justifiable method of estimating the overall 5-minute capsize probability using model test results for a single-wave condition. That is a single combination of *Hs* and *Tz*. Such a method can never be rigorously linked with the safety objective, but it is proposed that it may be regarded as a conservative approximation. (b) Test methodology The proposed test methodology is as follows: The rotorcraft designer selects a desired significant wave height limit *Hs(limit)* for ditching or the emergency flotation certification of his helicopter. Model tests are performed in the sea condition *Hs(limit) Tz(limit)* (where *Tz(limit)* is the zero-crossing period most likely to accompany *Hs(limit)*) with the selected spectrum shape using the method specified above, and the 5-minute probability of capsizing (*Pcapsize*) established in this sea condition. The way in which *Pcapsize* varies for other values of *Hs* and *Tz* is not known because it is not proposed to perform model tests in all the other possible combinations. Furthermore, there is no theoretical method to translate a probability of capsizing from one sea condition to another. However, it is known that the probability of capsizing is related to the exposure to breaking waves of sufficient height, and that this is in turn linked with wave steepness. Hence: (1) the probability of capsizing is likely to be higher for wave heights just less than Hs(limit) but with wave periods shorter than Tz(limit); and (2) the probability of capsizing will be lower for the larger population of wave conditions with wave heights less than Hs(limit) and with wave periods longer than Tz(limit). So, a reasonable and conservative assumption is that on average, the same Pcapsize holds good for all wave conditions with heights less than or equal to Hs(limit). A further conservative assumption is that Pcapsize is unity for all wave heights greater than Hs(limit). Using these assumptions, a comparison of the measured Pcapsize in Hs(limit) Tz(limit) against the target probability of capsizing (Pcapsize(target)) can be performed. In jurisdictions where flying is not permitted when the wave height is above Hs(limit), the rotorcraft will have passed the certification criteria provided that Pcapsize ≤ Pcapsize(target). In jurisdictions where flying over waves greater than *Hs(limit)* is permitted, the rotorcraft will have passed the certification criteria provided that: *Pcapsize**≤ Pcapsize(target)**– Pe*, where *Pe* is the probability of exceedance of *Hs(limit)*. Clearly, in this case, it can be seen that it would not be permissible for the rotorcraft designer to select an *Hs(limit)* which has a probability of exceedance greater than *Pcapsize(target)*.  Figure 1 — Example of all-year wave scatter table [Amdt No: 27/5]