Air to Air Refuelling Simulation:
The Hunter Hose & Drogue Model 

Hose and Drogue refuelling: a weakness in AAR simulation

 

A key aspect of NATO interoperability is the sharing of Air to Air Refuelling (AAR) tanker assets, so pilots need to be able to operate with and receive fuel from a variety of tankers. Simulation training is typically used to achieve operational familiarity with tankers from friendly nations before deployment.

​

There are two main methods of AAR operation: Boom, and Hose and Drogue. With Boom, an operator on the tanker 'flies' a rigid boom to contact with a receiver aircraft. This method is predominantly used by the USAF. With Boom it is easier to make contact, but harder to maintain it. 

​

With Hose and Drogue, the receiver pilot flies to contact with a drogue at the end of a hose trailed by the tanker. With Hose and Drogue it is harder to make contact, but easier to maintain it once connected due to the flexibility of the hose. Hose and Drogue is the most common method of AAR and is used by the USMC and most Western European aircraft.

​

Boom AAR operations are reasonably well simulated, but a weakness in AAR simulation training is the receiver flying to contact with hose and drogue. This is one of the most demanding pilot handling tasks and one that is particularly challenging to simulate. Until now, the lack of fidelity of real-time AAR hose and drogue simulation has meant that AAR receiver simulator training is only procedural at best. This means flying to AAR contact is one of the few remaining high workload pilot tasks that has to be largely trained live in the aircraft.

​

Live flying AAR is expensive, requiring by definition two flying aircraft – and often a third for a flying instructor if no two-seat trainer is available. It is also particularly hazardous due to the enforced close proximity with the potential to damage drogues, probes and pitot tubes. Even FOD being ingested into engines is not uncommon. With 5th and 6th generation fighters, the introduction of composite materials and stealth paint schemes means that contact between drogues and the receiver airframe is now far more costly than a traditional light scratch of paint.

​

In addition, due to the inherent risks involved, it is not possible to train pilots during live flying to deal with fairly common "failure" scenarios such as dead hoses, damaged drogues or mild turbulence.

​

The problems with the traditional hose and drogue AAR modelling approach

​

Many elements need to be modelled for hose and drogue AAR, but they can be summarised as follows:

​

1. Modelling the tanker: flight test data from the aircraft manufacturer is required for tanker flight dynamics, refuelling operator and tanker fuel system, and tanker wake vortex effects to be modelled realistically requires flight test data from the aircraft manufacturer.

2. Modelling the receiver: flight test data from the aircraft manufacturer is required for the receiver aircraft flight dynamics and fuel system to be modelled realistically. 

3. Modelling the probe, hose and drogue: the required data from the AAR manufacturer is generally not available to flight simulator manufacturers for the hose drum unit, hose aerodynamics and structural characteristics, drogue and coupling aerodynamics and structural characteristics, and receiver bow wave and probe to drogue contact. 

​

For both the tanker and receiver models, the limiting factor in simulation is the availability of good quality data. Where such data has been available, generally, good quality models have been produced. This is not where the modelling gap exists that currently limits AAR training, at least for fixed wing simulations. 

​

For hose and drogue pilot training simulation, modellers have typically resorted to effects-based modelling; for example, approach off centre and you will push the drogue away, approach at the correct closure rate near the centre of the drogue and you will make contact. Such simplifications produce erroneous behaviour in the many possible configurations of probe and drogue interaction. 

​

High fidelity, non-real-time physics based models have long been used by AAR equipment manufacturers for hose and drogue to predict performance of these non-linear systems. As computing power has increased, these complex models have been able to run in near real time. However, there are inherent difficulties on a flight simulator in adapting models not specifically designed to run in real time, especially in the critical contact phase.

​

What has been needed is a high fidelity model that realistically simulates the behaviour of the hose and drogue in real time with a piloted aircraft in the challenging probe/drogue contact regime. This would considerably improve the training capability and provide enhanced training transfer and operational capability with reduced cost of operations and reduction in environmental impact.

​

How HSL tackled the hose and drogue modelling challenge

​

Improving the hose and drogue modelling deficiencies and closing the modelling gap has been made possible by several convergent factors.

​

Those same NATO standards that enable interoperability of nations define the data needed to model the AAR equipment dimensions and modes of operation. In addition, HSL was aware of work being done at the Aerial Refuelling Systems Advisory Group (ARSAG) to define an ICD for modelling and simulation of AAR. 

​

At the same time, a beta release of Matlab-Simulink became available with multi-body physics modelling features.

​

Finally, a happy coalescence of available highly skilled engineers became available at the same time as HSL had sufficient in-house development funding available and a pandemic-induced lull in work. The team had exactly the right experience of flight simulation, AAR pilot training simulation and AAR simulation and modelling.

​

Beginning with a clean sheet, HSL launched a six-week proof of concept activity, focussed on the risk areas of contact modelling and real time execution of Matlab-generated code at high enough execution rates for stability. The proof of concept met the design targets and with a subsequent initial contract award, development commenced on the HSL Hose and Drogue model. 

​

Initial deliveries were made after 6-8 months, with continuing development and fine-tuning over the next year. The benefits of physics modelling were seen when the dead hose and damage drogue behaviours fell naturally from the model, with no need for special effects.

​

The initial releases were highly praised and the most recent model release with such effects as sliding to contact down the inner parts of drogue has been acclaimed as the best simulation that the test pilots have seen. 

​

Lessons Learned

 

Subject matter expertise was required in both pilot training AAR flight simulation and AAR equipment modelling and simulation. Without one the model behaviour would not be sufficiently realistic, and without the other the model would not run in real time properly correlated with the rest of the simulation. Matlab has some very good features but without good SME knowledge and tool training it is far too easy to produce an impressive looking model with little idea of what the model is actually doing. Processing power is no longer a limiting factor and a high fidelity model using latest techniques can be run stably at rates as low as 200Hz.

​

Correct order of execution and synchronous computation of tanker, receiver, hose and drogue and visuals with precisely correlated positions and dimensions is still essential, as stability and fidelity are highly sensitive to small errors.

​

Future activities include feedback to update the ARSAG ICD, switchable hose and drogue model fidelity levels for ab initio, recurrent training and CGF receivers, and assessment of objective fidelity measurement criteria such as video matching.

​

In Conclusion

 

Using openly available data and standards together with flight simulation and AAR subject matter expertise and the latest tools and techniques, HSL has now closed the AAR hose and drogue modelling gap and enabled significantly enhanced training transfer.

​

Within the HSL-led team, we can support you in any endeavour that requires high fidelity hose and drogue modelling. Talk to us today, and see how we can help.

​

Authors:

• Chris Hunter CEng FRAeS, Company Director at HSL, has 35 years of flight simulation experience including as Head of Engineering at Thales Training & Sim and L-3 Link UK, Chief Engineer on A400M STE and Nimrod MRA4 and committee member and past chairman of the RAeS Flight Simulation Specialist Group.

• Henry Clarke IMechE, Consultant at BadgerWorks, has 20 years experience in aerospace fuel systems, including air to air refuelling systems development, and the modelling of such systems at Cobham where he was responsible for AAR system performance and held the posts of Engineering Manager of Simulation & Modelling & Department and Senior Engineering Manager Research & Development.