Final Report Summary - EXTICE (EXtreme ICing Environment)
Recent aircraft incidents and accidents have highlighted the existence of icing cloud characteristics beyond the actual certification envelope currently defined by the specifications and federal aviation regulations (CS/FAR) Appendix C, which accounts for an icing envelope characterised by water droplet diameters up to 50 µm, the so-called cloud droplet. The main concern is the presence of super-cooled large droplets (SLD) such as freezing drizzle, in the MVD range of 20 to 110 µm, or freezing rain, with maximum droplet diameter beyond 500 µm. The presence of SLD was also confirmed in Europe by the European funded project EURICE. The main results raised within the EURICE project were that, while the existence of SLD was proved, means of compliance and engineering tools to accurately simulate these conditions were lacking and existing measures should be improved and new techniques should be developed.
International airworthiness authorities, namely the Federal Aviation Administration (FAA), Transport Canada (TC), and the European Aviation Safety Agency (EASA) are intending to jointly develop and issue updated regulations for certification in SLD. The outcomes of such an international effort to improve flight safety are a comprehensive proposal for a brand new set of regulations known among insiders as 'Appendix O'.
Once implemented, the proposed new rules will require aircraft manufacturers to demonstrate that their product can safely operate in SLD environments. To do so, they will be requested to demonstrate that specific capabilities comply with the new regulation. Compliance has typically involved actual flights into natural icing conditions, as well as the use of engineering simulations of the natural environment provided by experimental means, icing tunnels and tankers and analytical methods, namely ice prediction computer codes.
Preliminary analyses of the existing tools have demonstrated the limitations of the ice modelling capabilities, particularly since these methods have focussed towards representing the Appendix C icing envelope and were not developed to account for SLD conditions. Since SLD icing conditions have much larger droplet sizes than Appendix C conditions, the current methods do not adequately represent the droplet dynamics and icing physics associated with SLD.
At the present time, certification authorities rely primarily on flight test data for icing certification. Unfortunately flight tests in icing conditions are costly and difficult to achieve. If standard icing conditions are not easy to meet during an icing flight campaign, flight test in extreme icing conditions, such as SLD conditions, are still more troublesome. Advantages over the present situation could be achieved by performing part of the certification process through a combination of wind tunnel testing and numerical simulations; however these approaches must be proven reliable and trustworthy. Indeed, to cover the SLD envelope, there exists a need to extend and improve existing wind tunnel techniques and numerical simulation tools.
The objectives of EXTICE project were twofold. The first objective was to reduce aircraft development cost by improving tools and methods for aircraft design and certification in an icing environment. The second objective was to improve safety by providing more reliable icing simulation tools.
In an effort to improve the reliability of simulations and to prove their accuracy, the methodology chosen integrated basic experiments, wind tunnel testing and flight testing.
Basic experiments were performed to improve SLD physics knowledge and their results were used to define single SLD droplet basic mathematical model that could be implemented in ice accretion numerical simulation tools.
Icing wind tunnel tests on 'industrial components' such as a wing or an airfoil were used both to validate and improve numerical tools by identifying the best approach to be used for ice accretion simulation. Finally, it was planned, within the EXTICE project, to evaluate ice accretion accumulated on a specific test article installed on an aircraft flying in icing conditions.
The most important achievements from the EXTICE project were the following:
1. a website was created to present the results of the EXTICE project and contributed significantly to the goal of disseminating information gathered during the course of the project activities.
2. a comprehensive literature review of the state of knowledge on SLD was completed and a report was prepared. This review provided a comprehensive summary of SLD cloud conditions and appropriate ranges for various parameters to be used during subsequent experiments.
3. the basic studies of SLD physics were suitably divided between small, medium-sized and large facilities, thereby minimising the overlap in research activities. Basic SLD related experiments, such as splash and breakup, were completed and added significantly to the pre-existing knowledge base, being of direct relevance to aircraft icing, e.g. splash with surface roughness present and splash directions covering the full azimuth, rather than just in the forward direction, which was not available from previous droplet and wall collision experimental data.
4. empirical expressions were determined for the deposited mass ratio. A deterministic model of splashing up to the point of breakup was developed and a new SLD splashing model was formulated for wet surface drop impact. Thus, these studies of the basic SLD physics significantly increased the knowledge of these phenomena.
5. all key experimental programs were completed, including the two-dimensional icing wind tunnel test in DGA, the three-dimensional icing wind tunnel test in CIRA IWT and the flight test programme that was conducted in March 2012.
6. a wide range of droplet trajectory and ice shape prediction codes were successively modified by various partners. In particular, both drop splashing and rebound phenomena were incorporated in the various codes used by the participating organisations. Some organisations also used different formulations for these phenomena and inter-compared them to determine which of them seemed to lead to the best match with the experiments.
7. comparisons between numerical simulation and two-dimensional data obtained in the French DGA, formerly CEPr, icing facility were successfully completed. Both Lagrangian and Eulerian methodologies for calculating drop trajectories were employed in EXTICE with similar results. It was noted that the Eulerian method showed good promise, especially when coupled with an Euler flow code.
8. the biggest improvement between code predictions and experimental results occurred for the cases with large MVDs when splashing algorithms were employed; however, the effect of SLD modifications in the various codes appeared to be small in some cases, but in other cases showed a significant improvement. At higher Mach numbers, the models failed to accurately predict leading edge ice accretion amounts and the losses due to drop splashing seemed to be underestimated. This was particularly true at warm temperatures and it was not known whether this was a SLD specific feature or a general limitation in the icing prediction codes. This suggested that further work on ice accretion code modelling was required in the future.
9. the three-dimensional experimental data (CIRA tests) were only gathered during the final stages of the project and only a partial numerical and experimental validation was performed. A code comparison against a subset of core cases was successfully achieved. All codes predicted a similar large difference in ice thickness with respect to experimental data for one run but were fairly good on all other cases. It was underlined that one critical test condition, of MDV equal to 165 µm had to be addressed due to underestimates of drop size that implied an incorrect evaluation of the LWC at high diameters with existing instruments. This condition proved that the limitation of the instrumentation for SLD facility calibration was one of the key aspects that needed to be addressed.
10. while ice shape predictions provided good results for the cold cases, at warmer temperatures none of the utilised codes were able to demonstrate an ability to predict the feathers and scallops observed in the experiments.
11. it appeared that the application of two-dimensional techniques to the three-dimensional cases were relatively successful, leading to the conclusion that the three-dimensional experiments did not lead to strongly three-dimensional effects except on the trailing edge of the extended slat.
12. the organisations running the experimental facilities were amongst the first in the world to implement capabilities to simulate the bimodal drop size distributions which were observed in previous flight campaigns. Furthermore, the experiments which were performed were able to quantify the importance of this capability.
The top level expected result was a fundamental knowledge of the SLD ice accretion environment analysis and the development of European theoretical and experimental capabilities, the so-called engineering tools, to accurately model SLD encounter effects on aircraft in order to comply with the coming new icing certification rules. At the same time a deeper knowledge of SLD impact on aircraft was obtained and countermeasures from SLD conditions were investigated.
In conclusion, the EXTICE project allowed for a decrease in time and costs for aircraft design and certification, as well as for an increase in aircraft safety by providing the industries with more reliable icing simulation tools. More specific information could be obtained at the project website 'extice.cira.it' or after contact with Giuseppe Mingione, CIRA, via Maiorise, 81043, Capua, Italy, telephone number +39 0823623614, and e-mail address 'g.mingione@ira.it'.
International airworthiness authorities, namely the Federal Aviation Administration (FAA), Transport Canada (TC), and the European Aviation Safety Agency (EASA) are intending to jointly develop and issue updated regulations for certification in SLD. The outcomes of such an international effort to improve flight safety are a comprehensive proposal for a brand new set of regulations known among insiders as 'Appendix O'.
Once implemented, the proposed new rules will require aircraft manufacturers to demonstrate that their product can safely operate in SLD environments. To do so, they will be requested to demonstrate that specific capabilities comply with the new regulation. Compliance has typically involved actual flights into natural icing conditions, as well as the use of engineering simulations of the natural environment provided by experimental means, icing tunnels and tankers and analytical methods, namely ice prediction computer codes.
Preliminary analyses of the existing tools have demonstrated the limitations of the ice modelling capabilities, particularly since these methods have focussed towards representing the Appendix C icing envelope and were not developed to account for SLD conditions. Since SLD icing conditions have much larger droplet sizes than Appendix C conditions, the current methods do not adequately represent the droplet dynamics and icing physics associated with SLD.
At the present time, certification authorities rely primarily on flight test data for icing certification. Unfortunately flight tests in icing conditions are costly and difficult to achieve. If standard icing conditions are not easy to meet during an icing flight campaign, flight test in extreme icing conditions, such as SLD conditions, are still more troublesome. Advantages over the present situation could be achieved by performing part of the certification process through a combination of wind tunnel testing and numerical simulations; however these approaches must be proven reliable and trustworthy. Indeed, to cover the SLD envelope, there exists a need to extend and improve existing wind tunnel techniques and numerical simulation tools.
The objectives of EXTICE project were twofold. The first objective was to reduce aircraft development cost by improving tools and methods for aircraft design and certification in an icing environment. The second objective was to improve safety by providing more reliable icing simulation tools.
In an effort to improve the reliability of simulations and to prove their accuracy, the methodology chosen integrated basic experiments, wind tunnel testing and flight testing.
Basic experiments were performed to improve SLD physics knowledge and their results were used to define single SLD droplet basic mathematical model that could be implemented in ice accretion numerical simulation tools.
Icing wind tunnel tests on 'industrial components' such as a wing or an airfoil were used both to validate and improve numerical tools by identifying the best approach to be used for ice accretion simulation. Finally, it was planned, within the EXTICE project, to evaluate ice accretion accumulated on a specific test article installed on an aircraft flying in icing conditions.
The most important achievements from the EXTICE project were the following:
1. a website was created to present the results of the EXTICE project and contributed significantly to the goal of disseminating information gathered during the course of the project activities.
2. a comprehensive literature review of the state of knowledge on SLD was completed and a report was prepared. This review provided a comprehensive summary of SLD cloud conditions and appropriate ranges for various parameters to be used during subsequent experiments.
3. the basic studies of SLD physics were suitably divided between small, medium-sized and large facilities, thereby minimising the overlap in research activities. Basic SLD related experiments, such as splash and breakup, were completed and added significantly to the pre-existing knowledge base, being of direct relevance to aircraft icing, e.g. splash with surface roughness present and splash directions covering the full azimuth, rather than just in the forward direction, which was not available from previous droplet and wall collision experimental data.
4. empirical expressions were determined for the deposited mass ratio. A deterministic model of splashing up to the point of breakup was developed and a new SLD splashing model was formulated for wet surface drop impact. Thus, these studies of the basic SLD physics significantly increased the knowledge of these phenomena.
5. all key experimental programs were completed, including the two-dimensional icing wind tunnel test in DGA, the three-dimensional icing wind tunnel test in CIRA IWT and the flight test programme that was conducted in March 2012.
6. a wide range of droplet trajectory and ice shape prediction codes were successively modified by various partners. In particular, both drop splashing and rebound phenomena were incorporated in the various codes used by the participating organisations. Some organisations also used different formulations for these phenomena and inter-compared them to determine which of them seemed to lead to the best match with the experiments.
7. comparisons between numerical simulation and two-dimensional data obtained in the French DGA, formerly CEPr, icing facility were successfully completed. Both Lagrangian and Eulerian methodologies for calculating drop trajectories were employed in EXTICE with similar results. It was noted that the Eulerian method showed good promise, especially when coupled with an Euler flow code.
8. the biggest improvement between code predictions and experimental results occurred for the cases with large MVDs when splashing algorithms were employed; however, the effect of SLD modifications in the various codes appeared to be small in some cases, but in other cases showed a significant improvement. At higher Mach numbers, the models failed to accurately predict leading edge ice accretion amounts and the losses due to drop splashing seemed to be underestimated. This was particularly true at warm temperatures and it was not known whether this was a SLD specific feature or a general limitation in the icing prediction codes. This suggested that further work on ice accretion code modelling was required in the future.
9. the three-dimensional experimental data (CIRA tests) were only gathered during the final stages of the project and only a partial numerical and experimental validation was performed. A code comparison against a subset of core cases was successfully achieved. All codes predicted a similar large difference in ice thickness with respect to experimental data for one run but were fairly good on all other cases. It was underlined that one critical test condition, of MDV equal to 165 µm had to be addressed due to underestimates of drop size that implied an incorrect evaluation of the LWC at high diameters with existing instruments. This condition proved that the limitation of the instrumentation for SLD facility calibration was one of the key aspects that needed to be addressed.
10. while ice shape predictions provided good results for the cold cases, at warmer temperatures none of the utilised codes were able to demonstrate an ability to predict the feathers and scallops observed in the experiments.
11. it appeared that the application of two-dimensional techniques to the three-dimensional cases were relatively successful, leading to the conclusion that the three-dimensional experiments did not lead to strongly three-dimensional effects except on the trailing edge of the extended slat.
12. the organisations running the experimental facilities were amongst the first in the world to implement capabilities to simulate the bimodal drop size distributions which were observed in previous flight campaigns. Furthermore, the experiments which were performed were able to quantify the importance of this capability.
The top level expected result was a fundamental knowledge of the SLD ice accretion environment analysis and the development of European theoretical and experimental capabilities, the so-called engineering tools, to accurately model SLD encounter effects on aircraft in order to comply with the coming new icing certification rules. At the same time a deeper knowledge of SLD impact on aircraft was obtained and countermeasures from SLD conditions were investigated.
In conclusion, the EXTICE project allowed for a decrease in time and costs for aircraft design and certification, as well as for an increase in aircraft safety by providing the industries with more reliable icing simulation tools. More specific information could be obtained at the project website 'extice.cira.it' or after contact with Giuseppe Mingione, CIRA, via Maiorise, 81043, Capua, Italy, telephone number +39 0823623614, and e-mail address 'g.mingione@ira.it'.