An electrical “robust channel” for aircraft hydraulic power generation. A design and test review

contenu protégé  Document accessible sous conditions - vous devez vous connecter ou vous enregistrer pour accéder à ou acquérir ce document.
- Accès libre pour les ayants-droit


An electrical “robust channel” for aircraft hydraulic power  generation. A design and test review


application/pdf An electrical “robust channel” for aircraft hydraulic power generation. A design and test review Thomas Delphin, François Biais
Détails de l'article
contenu protégé  Document accessible sous conditions - vous devez vous connecter ou vous enregistrer pour accéder à ou acquérir ce document.
- Accès libre pour les ayants-droit

An electrical “robust channel” for aircraft hydraulic power  generation. A design and test review


139.97 Ko


Creative Commons Aucune (Tous droits réservés)





<resource  xmlns:xsi=""
        <identifier identifierType="DOI">10.23723/10638/20117</identifier><creators><creator><creatorName>François Biais</creatorName></creator><creator><creatorName>Thomas Delphin</creatorName></creator></creators><titles>
            <title>An electrical “robust channel” for aircraft hydraulic power  generation. A design and test review</title></titles>
        <resourceType resourceTypeGeneral="Text">Text</resourceType><dates>
	    <date dateType="Created">Sun 1 Oct 2017</date>
	    <date dateType="Updated">Sun 1 Oct 2017</date>
            <date dateType="Submitted">Thu 21 Mar 2019</date>
	    <alternateIdentifier alternateIdentifierType="bitstream">631300c46ee30bb905766538ff5df1a36cd3c1ba</alternateIdentifier>
            <description descriptionType="Abstract"></description>

An electrical “robust channel” for aircraft hydraulic power generation. A design and test review. Thomas DELPHIN (1, 2), François BIAIS (2) 1 : University of Toulouse, INPT, LAPLACE ; 2 rue Charles Camichel, BP 7122, 31071 Toulouse Cedex 7, France ; 2 : THALES AES ; 41 boulevard de la République, BP 53, 78401 Chatou Cedex, France ; {thomas.delphin ; francois.biais} Abstract In the frame of weight reduction studies for the More Electrical Aircraft, an electrically driven pump for “robust” hydraulic power generation has been developed by THALES AES within CLEANSKY program in collaboration with AIRBUS. This concept called “robust channel” allows removal from the engine of conventional engine driven hydraulic pumps, but requires reliable and robust operation due to critical hydraulic loads such as flight controls and landing gear operations. This study is based on the direct drive of a pump Induction Motor (IM) by a constant U/f variable frequency network supplied by a Permanent Magnet Generator (PMG), without the need of intermediate power conversion. However the absence of field control of the PMG and its impedance have a significant impact on the network voltage, particularly during direct starting of the pump because of the IM inrush current. The inherent interdependency of both machines led to design a channel as a whole, which was built and tested in many operating conditions. Introduction Conventionally, the aircraft hydraulic pumps are mechanically driven by the gearbox, the speed of the pump being linked to the speed of the engine shaft. Some alternate electrical solutions are based on an electrical pump motor driven by a power converter supplied by the aircraft constant voltage network [1]. A non-conventional solution is presented here, a variable speed variable frequency variable voltage specific channel, consisting of a Permanent Magnet Generator (PMG) driven by the gearbox and powering directly without power converter an Induction Machine (IM). However, while this non-conventional solution offers attracting advantages, such a wild frequency and wild voltage channel is a design challenge. Design considerations A PMG offers a very efficient solution to reduce the generator weight; its well-known advantages compared to conventional 3-stage generator are lower weight, simplicity, robustness and efficiency. The weak point is obviously the absence of field control, which prevents voltage regulation and field extinction. When driven by the gearbox, because of its constant field, the PMG naturally provides a variable voltage U at a variable frequency f, both U and f proportional to the speed, at a constant U/f ratio; this feature is suitable to the use of an IM driving the pump because a constant U/f means a constant flux within the IM, which can therefore operate optimally in the whole frequency range. In addition, the direct start of the IM in the whole frequency range, which is the most difficult requirement to meet, is facilitated by the theoretical constant U/f. The starting torque produced by the IM is proportional to U 2 /f 3 ; therefore at a constant U/f instead of a constant U as would be a conventional supply, this torque is much higher. Nevertheless one must take into account the impact of the non-regulated voltage of the PMG: when loaded and particularly during the high inrush current of the IM, because of the impedance of the PMG there is a significant voltage drop at its terminals [2] [3]; this will affect the operation of the IM. In addition, since this robust channel concept is related to a specific channel, the PMG is more or less a relatively low power generator, and as a consequence has a high internal impedance. This will have a direct impact on the starting torque capability of the pump IM because its starting torque is driven by the square of the supply voltage. Since reducing the impedance of the PMG means increasing its power rating (and its weight), it is necessary to design this channel as a whole to avoid oversizing and find an optimal weight of the system. On the IM side, the starting torque capability is dependent on its rotor resistance and leakage inductance [4], including frequency effects for cage design suitable to direct start (deep cage, double cage). The impact of other potential loads connected to the channel needs also to be addressed if relevant, because of an additional associated voltage drop in the network. During the study of this channel, a steady state model has been used at first, Fig. 1, with the aim of identifying the key parameters and finding a reference solution meeting the starting torque requirements. Fig. 2 shows an example of the impact of the PMG impedance on the IM torque-speed characteristics and particularly the starting torque: in this case, a PMG impedance increase by 50% produces an increase of the voltage drop caused by the IM inrush current, which leads to a starting torque reduction of approximately 25%. Fig. 1: Steady state model of the robust channel Fig. 2: Impact at 800 Hz of PMG inductance, L, on IM torque-speed characteristics; L (black), 1.5L (red). Based on a preliminary requirement specification on pump performance (rating and starting) and additional load on a 400-800 Hz channel, a PMG and IM system was designed with a set of following characteristics: PMG impedance and no-load voltage, IM rotor resistance and leakage inductance. Both machines incorporating these characteristics were built, tested separately to validate their parameters, and then tested as a system so as to validate the steady state model used. Testing of the “robust channel” Many test sequences, including very specific cases, were investigated in the operating frequency range [400Hz - 800Hz]. The most critical sequences are presented in this paper and listed in the table hereunder. Test sequence 1 Direct on line start at 800Hz on load Test sequence 2 Power break Table 1: List of the test sequences presented Test sequence 1 illustrates the main function of the ”robust channel”. It shows the capability of the IM to start under the most difficult conditions, i.e. when loaded and supplied at high frequency. Test sequence 2 is also investigated because the transient effects can be very stressful for the system [5] [6]. In the following figures the steady state curves computed with the measured parameters of the IM and PMG (including frequency effect and main flux path saturation) are plotted in order to show the steady state model accuracy. Since the absolute value of sinusoidal functions is plotted, only the envelop of the curve is to be compared with the steady state curve. Direct start test sequence The mechanical model is not in the scope of this article therefore the dynamic of the IM is not accurately modeled as it is shown in Fig. 3. Indeed, the slope of the speed computed with the steady state model does not fit speed recorded. Thus the time to reach the steady state is not accurate in all the results presented in the following figures. In Fig. 3, we can see that the model fails to predict the torque oscillations at the beginning of the starting. Also, the mean value of the torque during the starting up does not fit because the leakage flux path saturation is not taken into account. However the peak torque magnitude seems accurate because at this instant this effect is reduced. This simplification of the model also leads to a significant error on the inrush current, Fig. 4, and consequently on the voltage sag, Fig. 5. It can be identified that the magnitude of the torque oscillation is 2.8 times higher than the nominal torque. Also the magnitude of the inrush current peak is 8 times higher than the nominal current. Fig. 3: IM Torque (N.m) versus time (s); simulation result (red), test result (light blue) Fig. 4: Absolute value of current (A) versus time (s); simulation result (red), test result (light blue) 0 2000 4000 6000 8000 10000 12000 14000 16000 0 5 10 15 20 25 30 35 40 45 50 rpm 0 0.5 1 1.5 2 2.5 3 -50 -40 -30 -20 -10 0 10 20 30 40 50 measured torque (Nm) versus time (sec) 0 0.5 1 1.5 2 2.5 3 0 50 100 150 200 250 300 350 400 Fig. 5: Absolute value of voltage (V) versus time (s); simulation result (red), test result (light blue) Specific transient test sequence We will now focus on specific transient problems. In the literature, power breaks have been identified as very stressful for IM, especially when the flux is still trapped in the airgap [6]. This configuration has been tested in the sequence 2. First, the loaded IM reaches the steady state at 400 Hz; the line contactor is then opened (torque and current at 0) and re-closed. In Fig. 6 and 7, we can see the three steps of this test sequence. Fig. 8 shows that there is still flux in the airgap when the line contactor is re-closed because the voltage is not 0 at this instant. In this case the transient torque magnitude is 4.8 times higher than the nominal torque and the current magnitude 10 times higher. Fig. 6: IM Torque (N.m) versus time (s) Fig. 7: IM 3 phase currents (A) versus time (s) Fig. 8: IM voltage (V) versus time (s) The results show that this operating condition is more stressful than the starting of the IM. Concerning the mechanical parts, the torque magnitude exceeds the expected peak value of the torque-speed characteristic. As for the electrical parts, the inrush current is much higher than in regular starting operation. These tests have brought to light that beyond the steady state model used to design the robust channel as a whole, a transient model is necessary during the design stage to predict such effects that can be constraining. [7]. Conclusions In this paper a non-conventional solution for an electrical pump drive has been presented. A PMG connected to the gearbox is supplying an IM at constant U/f without regulation. However the balance between the impedance of the PMG and the inrush current of the IM is a difficult problem to guarantee the performances of the pump. Based on a steady state model of the “robust channel”, a PMG has been built and tested. Tests revealed that the “robust channel” is functioning in all configurations. However, results showed that some transient effects such as torque oscillations during the starting up of the IM are not modeled. Moreover, some specific transient sequences underlined constraining operating modes that should be investigated during the design stage. As a consequence a transient model is being developed to complete the steady state one. References 1 A. A. AbdElhafez et al, “A Review of More-Electric Aircraft”, 13th International Conference on Aerospace Sciences & Aviation Technology, 2009 2 X. Liang et al, “Induction Motor Starting in Practical Industrial Applications”, IEEE Transactions on Industry Applications, 2011, Vol.47, pp.271-280 3 E. Tuinman et al, “Simulation of a ‘direct on line’ start of a large induction motor connected to a salient pole synchronous generator”, International Conference on Simulation, 1998, Vol.1, pp.362-367 0 0.5 1 1.5 2 2.5 3 0 50 100 150 200 250 300 350 400 1.55 1.6 1.65 1.7 1.75 1.8 -60 -40 -20 0 20 40 1.55 1.6 1.65 1.7 1.75 1.8 -500 -400 -300 -200 -100 0 100 200 300 1.55 1.6 1.65 1.7 1.75 1.8 -150 -100 -50 0 50 100 150 4 C. H. Lee, “A Design Method for Double Squirrel- Cage Induction Motors”, Power Apparatus and Systems, Part III. Transactions of the American Institute of Electrical Engineers, 1953, Vol.72, pp.630-636 5 H. G. Tempelaar, “Determination of transient overvoltages caused by switching of high voltage motors”, IEEE Transactions on Energy Conversion, 1988, Vol.3, pp.806-814 6 M. Baran et al, “Stresses on Induction Motors Due to Momentary Service Interruptions”, 2006 IEEE Industrial and Commercial Power Systems Technical Conference, 2006, Vol.1, pp.1-8 7 A. H. Bonnet, “Root Cause AC Motor Failure Analysis with a Focus on Shaft Failures”, IEEE Transactions on Industry Applications, 2000, Vol.36, pp.1435-1448