How do you know if you need to use a safety abort manifold? What options do you have in its selection?
Hydraulic actuators provide a long life, economical, and low maintenance solution to applying load for the durability and ultimate strength testing of aircraft. However the high cost of test articles (in many cases a whole aircraft) and the long duration of fatigue tests, means that the probability of accidental damage to the test article due to equipment malfunction has to be minimized to (almost) zero. Consider that there is no possibility of carrying out a second test if an initial test is ruined. This means that systems designed need to tolerate component failures without compromising the safety of the structure under test. The problem is increased dramatically in situations where there can be more than one hundred actuators, mostly providing force control, simultaneously applying load to a complex structure that has a high level of cross coupling between channels.
A proven safety-by-design approach to this problem is to assume that every component (electrical and mechanical) can fail one at a time, possibly in multiple ways, and that there is hydraulics hardware, electronic controllers and control software capable of dealing safely with these failures. Examples of failures might include: solenoid valves failing to turn on as well as fail to turn off, servovalves failing in full-flow or a no-flow conditions, and relay contacts or semiconductor switches fused closed as well as open.
To provide an adequately safe system, a significant number of hydraulic components are required to supplement the primary servovalve, hence the need for a manifold as the smallest and most cost effective packaging solution for the hydraulics components. The following provides some of the reasoning for combining the main components within a safety abort manifold, and an insight into what influences the selection.
To achieve the desired control response, an optimally-sized actuator and servovalve will have full scale load pressure about two thirds of the supply pressure. This could mean an incorrectly configured controller could accidentally apply 50% overload, which could completely ruin the fatigue specimen. To avoid this, tension and compression cross-port relief valves are needed across each (equal area) actuator to limit the load independent of software settings. The limits are set typically 10% above the test full-scale load. In addition, to supplement this and cover several other possible single component failures, it is standard practice to have dual load cells on all loading channels to provide independent redundant load limit sensing and shutdown within the controller.
To accommodate a failure, the servovalve needs to be isolated from the actuator to either prevent further oil getting to the actuator or to prevent uncontrolled unpredictable leakage from the actuator through the servovalve spool. To provide this isolation, a pair of pilot-operated servo lock valves are needed between the servovalve and the actuator.
If some actuators have leakage, either by seal wear or deliberately to reduce friction, then having the system remain in a locked up state is undesirable. This is because it is possible because of structural interaction for the load in some locked up actuators to increase in load as the load is shed by other leaking actuators. For this reason it is desirable to quickly remove all load from the structure, typically over a period of five to ten seconds.
The unloading mechanism chosen for a system depends mainly on what the test structure can safely withstand under the test range of complex unloading conditions. In most cases, a Passive Abort is acceptable; however for this approach the time it takes to unload each actuator is dependent upon initial load in that actuator at the start of shut down. It is possible to adjust the needle valve for each actuator to have all actuators unload in the same time from one high-level complex load condition. However a shutdown from a different multi-channel load condition will result in some actuators unloading faster than others, which for some tests can cause undesirable reaction imbalances during the unload cycle. If however, balanced unloading under all possible initial conditions is required to maintain a reaction balance, then Controlled Abort is required. This can be Controlled Load Abort, Controlled Pressure Abort or even Controlled Displacement Abort depending on the feedback variable supplied. Any of these options have the advantage that they can be electrically set up. However it does add considerable cost because of the need for an additional unload control servovalve, transducers and an unload controller. It should be noted that the unload valve needs to be a Moog Direct Drive Valve as the pilot pressure will have been switched off or has failed during the unload cycle and this technology is ideal for this situation.
Although it is undesirable to have non-equal area actuators for control and cross-port relief reasons, it may be necessary. If there is no way of avoiding this type of actuator, then a more complex controlled unloading controller and unloading control valve configuration is required than is needed for equal area actuators. This is because to achieve the desired unload waveform, pressure has to be sometimes removed from the chamber with the lower pressure.
In conclusion, safety abort manifolds are essential on fatigue and ultimate strength tests where accidental damage to the structure under test cannot be tolerated. The decision to use a Passive Abort Manifold rather than Controlled Abort Manifold depends on the budget, set-up convenience, structural limitations and test requirements. The Moog Abort Manifolds do have a common base manifold for all configuration options.
The Hawk Lead-In-Fighter Full-Scale-Fatigue-Test uses a both Moog’s Passive and Controlled Abort Manifolds under the control of an MTS Aero ST unload controller that is completely independent of the primary control system.
Author
Graeme William Burnett is a guest contributor for the Moog newsletter sharing his expertise gained in over 35 years of work experience with the Defence Science and Technology Organisation (DSTO), Department of Defence, Melbourne, Australia as an engineer in support of aeronautical research. He is currently the Engineering Manager, Hawk Lead in Fighter, Full Scale Fatigue Test contracted to DSTO, Melbourne, Australia.