Simulating HIRF and Lightning Effects
Electromagnetic environments as, for instance, high intensity radiated fields (HIRF) and lightning can degenerate aircraft safe operation and even cause catastrophic effects [1] without any precautions taken. The trend toward all-electric aircraft increases the number of aircraft functions performed by avionics. The risk of failure is even higher for modern aircraft with structures made of carbon-fiber composites (CFC).
Today, besides prototyping, the electromagnetic (EM) simulation is a well-accepted design and development method. This is especially true for antenna design. On the other hand, electromagnetic environmental effects (E3) and electromagnetic compatibility (EMC), respectively, have been for a long time considered as black magic. Testing has been the preferred method to achieve compliance with standards. However, EM simulations for EMC are now also used for certain applications, i.e. in electronic design [2]. In the author’s opinion, the usage of EM simulations as part of the aircraft certification with respect to electromagnetic environments as HIRF and lightning is still in the early stages.
Knowledge about the internal EM environment of an aircraft where the avionics and linking cables are situated is important to any aircraft manufacturer. This allows the risk of electromagnetic environmental effects on critical avionics to be analyzed and mitigated, reducing weight and, with it, manufacturing and operational cost. Unlike aircraft testing, the EM simulation allows the internal environment to be determined along the entire product life cycle, even before the first prototype is built.
In this article, a modified aircraft model based on Evektor’s EV55 Outback plane [3] exposed to a HIRF and lightning environment, respectively, is considered for 3D electromagnetic E3 simulations. Specialized EM simulation software [4], including static and full wave solvers and a cable solver, was used to solve the electromagnetic problems. Field coupling into the aircraft is considered as well as coupling into cables to determine the internal EM environment of the aircraft.
Aircraft Electromagnetic Environments
Standardized electromagnetic environments to which a civil aircraft may be exposed during normal operation are defined in guidance documents. The HIRF environment, its development process and related testing methodologies are described in SAE ARP5583A [5]. The lightning electromagnetic environment and related test waveforms are defined in SAE ARP 5412B [5].
The HIRF Environment
The HIRF environment is due to the radiation of radio frequency energy into free space by television stations, radar stations and other sources. It covers the frequency range from 10 kHz up to 18/40 GHz. The frequency range is further divided into 17 frequency bands. For each frequency band in Table 1, field strength values in term of a peak and average values are given.
|
Frequency |
Field Strength / (V/m) |
|
|
PEAK |
AVERAGE |
|
|
50 |
50 |
|
|
50 |
50 |
|
|
… |
… |
… |
|
12 GHz – 18 GHz |
2000 |
200 |
|
18 GHz – 40 GHz |
600 |
200 |
Table 1: Certification HIRF environment
The Lightning Environment and Waveforms
Aircraft encounter lightning strikes while airborne or stationary on a runway. While the later event is in general more severe, the former one is more frequent [6]. An aircraft typically encounters about one strike per year while airborne and only one every hundred years while stationary.
A lightning event consists of multiple phases as shown in Figure 1 for a downward cloud-to-ground strike. It is started by a corona discharge, which initiates a leader. This leader starts at a cloud and propagates towards ground where an aircraft may be located. While it’s approaching ground, the electric field between the leader’s tip and ground increases resulting in final air breakdown. This initiates a second leader moving upward, starting at the aircraft. Once the downward and upward leaders meet at some height, a low impedance path between the cloud and ground forms. This finally allows a strong return stroke current to flow [7].
The lightning event of an airborne aircraft also starts in general with a lightning leader at cloud charges. The lighting leader approaching the aircraft increases the electric field close to the aircraft resulting in corona breakdowns at the aircraft surface and lighting leaders starting at the aircraft. In Figure 2 a bi-directional leader at the aircraft nose and tail is shown. Once a low impedance path is formed (of which the aircraft is part) a return stroke current flows neutralizing the cloud charges.
Lightning strikes originate from charge centers in thunderclouds, in particular cumulonimbus clouds. Typically, positive charges are accumulated at the top of these clouds while negative ones remain at the lower part as shown in Figure 2 [8]. These electric charges produce electrostatic fields inside clouds, between clouds and between clouds and ground. When an airborne aircraft passes such an ambient electrostatic field the electric field will be locally enhanced, i.e. at aircraft extremities with small curvature radius as shown in Figure 3. These field enhancements finally initiate the former mentioned corona breakdowns at the aircraft itself.
Return stroke current waveforms vary from one lightning event to another. Typical waveforms and sequences are standardized and defined in SAE ARP 5412B. The main sequence is shown in Figure 4. It consists of four waveform components: A (first return stroke), B (intermediate current), C (continuing current) and D (subsequent return stroke).
To consider standardized waveforms in numerical simulations, their profile has to be described in closed form. In SAE ARP5412B, an idealized exponential standard waveform, which refers to all waveforms except C, is defined as:
i(t) = I0∙(e–αt – e–βt)∙(1 – e–γt)2 (1)
Table 2 summarizes the parameter values associated with (1) for each of the components in Figure 4.
|
Parameter |
Lightning Current Component |
|||
|
A |
B |
C |
D |
|
|
I0 / A |
218 810 |
11 300 |
400 |
109 405 |
|
α / s |
11 354 |
700 |
– |
22 708 |
|
β / s |
647 265 |
2000 |
– |
1 294 530 |
|
γ / s |
5 423 540 |
7 253 750 |
– |
10 847 100 |
Table 2: Lighting current component parameters
In Figure 5, the current waveform A is shown as function of time t in microseconds. It has a peak current of 200 kA, a time to peak of 6.4 μs and a time to half-value of 69 μs.
The frequency spectrum of component A is shown in Figure 6. Compared to the HIRF environment as given in Table I, the lightning environment has rather low frequency content up to only few MHz.
Virtual Aircraft
In Figure 7, the modified aircraft model based on Evektor’s EV-55 Outback airplane is shown. It has a wing span of 16.10 m, a length of 14.21 m, and a height of 5.13 m. For the purpose of the electromagnetic simulations described in this paper only its fuselage was used. It was imported into the EM simulation software from an IGES file and defined as an infinitely thin perfect electric conductor (PEC) sheet. Composite materials, in particular carbon-fiber composites (CFC), can also be defined.
A 13 m long RG 58 coaxial cable was placed inside the aircraft to also consider electromagnetic field coupling with cable currents. It consisted of a solid copper signal wire and a braided shield. The signal wire was terminated by 45 ohm loads on both cable ends and the shield was directly bonded to the aircraft skin. EM field probes and current probes, respectively, in Figure 8 were defined to calculate EM fields at points outside and inside the aircraft and cable currents of the coaxial cable.
Unlike a real prototype, a virtual aircraft can easily be placed in free space to mimic the airborne situation. To allow for the open space in the three dimensional electromagnetic simulation, the computational volume shown in Figure 9 was terminated by an open boundary condition (PML).
Finally, to solve the electromagnetic problem of the aircraft, hexahedral meshes as shown in Figure 10 were created.
Electromagnetic Simulation
The electromagnetic behavior of the modified aircraft in a HIRF and lightning environment, respectively, were solved with the specialized EM simulation software we used. The static and full wave solvers as well as the cable solver are fully integrated into the same graphical user interface, allowing the use of the same models with different solvers from DC up to very high frequencies.
The electrostatic solver available in the EM simulation software package was used to solve the electrostatic problem of the aircraft in an ambient electrostatic cloud field. To solve the time-dependent electromagnetic problems of the aircraft in a HIRF and lightning environment, respectively, the full wave time-domain transmission line matrix (TLM) solver coupled with the cable solver was used. This integration of both solvers is dedicated to considering the electromagnetic effects of complex cable harnesses in electromagnetic environments.
The Transmission Line Matrix (TLM) is best suited for large scale E3 applications as described in this paper. It offers a kind of conformal meshing and octree based meshing for accurate and fast computation of arbitrarily shaped objects [4]. This technology allows the real aircraft geometry to be represented in a relatively coarse mesh, and thus can significantly reduce the memory and computation time requirements compared to EM simulations of similar accuracy on staircase mesh. Furthermore, compact models for slots, seams, and CFC as frequency dependent thin panels among others can be used to avoid excessively fine meshes and thus very small time steps, compared to the duration of a lightning strike up to 0.5 ms. State-of-the art, high performance computing capabilities (multithreading, GPU, and MPI computing) allow problems for several billions of degrees of freedom to be solved in a reasonable time frame [10].
HIRF Simulations
In SAE ARP5583A, beside the HIRF environment, aircraft high level and aircraft low level coupling tests are described. High-level tests involve the illumination of aircraft with high amplitude RF fields as specified in Table I. This is to evaluate the electromagnetic environmental effects (E3) on aircraft systems. The preferred low level coupling tests use low amplitude RF fields to determine internal aircraft electromagnetic environments. Internal aircraft electromagnetic environments can also be numerically calculated by means of electromagnetic simulations [10].
The modified aircraft in Figure 11 is illuminated by a plane wave with an electric field of 1 V/m. The angle of incidence angle is θ = 135 degree and φ is changed from 180 to 270 degree in 45 degree steps.
The surface current at 1 GHz for a plane wave incident from φ = 225 degree is shown in Figure 12. Also visible is the bulk current on the cable shield.
Figures 13 and 14 show the induced currents on the coaxial cable in the frequency range from 100 to 500 MHz typical for a low level swept current (LLSC) test. The results for all angles of incidence φ are very similar and show repeated resonance peaks related to the cable length of 13 m.
The signal current amplitudes in Figure 14 are about factor 1000 lower than the bulk current amplitudes in Figure 13 due to the braided cable shield of the coaxial cable and the closed connection of the cable shield to the aircraft fuselage at both cable ends.
In Figure 15, the electric field strength at the internal field probe VTP3 in Figure 8 is shown over the frequency from 500 MHz to 1000 MHz. This frequency range is typical for a low level swept field (LLSF) test which even extends up to 18 / 40 GHz. In difference to the cable current results the electric field strength shows a stronger dependence on the plane wave position. The strongest field coupling is for broadside illumination (θ = 135, φ = 270 degree) of the aircraft, which is reasonable as the field couples easily through the side windows into the aircraft.
The LLSC and LLSF simulations were performed on up to 4 NVIDIA Tesla S2050 GPUs, with total simulation times of around 1 hour and 5 hours, respectively, per angle of incidence. The hexahedral meshes comprised about 6 million (LLSC) and 42 million (LLSF) mesh cells, corresponding to 36 million and 252 million degrees of freedom.
Lightning Simulations
As mentioned previously, most lightning events encountered by an aircraft encounters by the aircraft when airborne. The electrostatic field of cumulonimbus cloud configurations is locally enhanced especially at the aircraft surface, with small curvature radius. It is in these local field enhancement zones of an aircraft that initial lightning strike attachment occurs [9]. These potential initial attachment zones can also be numerically calculated with electrostatic simulations in the simulation software we used.
Figure 4 and Figure 16 show the electrostatic field amplitude in selected planes for an initial ambient field of 1 V/m. While in Figure 4 the ambient field pointing from nose to tail is aligned with the longitudinal aircraft axis in Figure 16 the ambient electrostatic is pointing off-axis. Both results show field enhancements of greater than factor 5 in especially at the aircraft extremities as nose and tail.
In Figure 17 a return stroke current is injected at the nose and extracted at the tail by conducting PEC wires. This models a configuration similar to Figure 2 once a low impedance path is established. The current loop is closed via the open boundary conditions in Figure 9.
In Figure 18 the surface current density at 1 kHz is shown over the entire aircraft skin when a current waveform A is injected into the aircraft.
The related cable currents are shown in Figure 19 as a function of time up to 500 μs.
Their amplitudes up to approximately 12A and 8mA, respectively, are several magnitudes below the injected current with a peak value of 200 kA. At the relative low frequencies of the waveform A, the aircraft fuselage still blocks the external field to penetrate into the aircraft as shown in Figure 20. Nevertheless, without the coaxial shield of the cable, the signal current would be significantly larger and in a range that might even damage avionics. This underlies the necessity of cable shields and the importance of EM simulations in terms of optimizing shielding effectiveness, weight and cost.
The electrostatic simulation in Figure 16 was performed on a laptop computer with Intel Core i7 processor and 16 GB RAM in about 9 minutes. Thereby the hexahedral mesh counted about 9 million mesh cells and 9 million degrees of freedom, respectively. The transient simulation up to 500 μs with a mesh of about 1.3 million mesh cells (7.8 million degrees of freedom) was performed on 2 NVIDIA Tesla K40 GPUs in about 5 hours.
Conclusions
3D EM HIRF and lightning simulations were successfully applied to the modified aircraft model of Evektor’s EV-55 Outback plane. The provided numerical results in terms of electrostatic field enhancements, surface currents – even on cable harnesses – internal aircraft EM environment and cable currents offer insight into the EM interaction of an aircraft with an electromagnetic environment. Often, these effects cannot be assessed by aircraft testing in the field, while the EM simulation allows investigating EM effects even before an aircraft is built. This means that state-of-the-art high performance capabilities allow large-scale simulations to be performed on real sized aircraft models up to the GHz-range in a reasonable time frame. 3D EM simulation can successfully be used to support future aircraft certification and complements the testing process.
Acknowledgements
The modified aircraft geometry of Evektor’s EV-55 has been available due to the courtesy and related work of EVEKTOR, spol. s.r.o. and the HIRF SE project consortium.
References
- M. Tooley, and D. Wyatt, Aircraft Electrical and Electronic Systems: Principles, Maintenance and Operation, Oxford, Butterworth Heinemann, 2009.
- www.microwavejournal.com/articles/23497-emc-simulation-in-the-design-flow-of-modern-electronics
- www.evektor.cz/en/ev-55-outback.aspx
- www.cst.com
- www.sae.org
- D. Morgan, C.J. Hardwick, S.J. Haigh, and A.J- Meakins, “The interaction with aircraft and the challenges of lightning testing,” Journal AerospaceLab, ONERA, Issue 5, Dec. 2012.
- www.erico.com/public/library/fep/strike/LT30373.pdf
- www.dtic.mil/dtic/tr/fulltext/u2/a222716.pdf
- H. W. Kasemir, “Static discharge and triggered lightning,” Proc. 8th International Aerospace and Ground Conference on Lightning and Static Electricity, Fort Worth, Tex., pp 24.1 – 24.11, June 1983.
- M. Kunze, Z. Reznicek, I. Muntenau, P. Tobola, and F. Wolfheimer, “Solving large lulti-scale problems in CST STUDIO SUITE,“ Proceedigns of ICEAA, Torino, pp. 110-113, 2011.
