Dish Verification Antennae (DVA)–1 demonstrates excellent performance at L-band and can operate reasonably up to 10 GHz. However, with recent technological advances, there is a push towards the development of high frequency radio telescopes up to Q-Band and more. As an attempt to demonstrate the capabilities of the composite radio telescopes at higher frequency range (up to Q-Band), a DVA–2 is currently under construction. In this article the authors will elaborate the design path towards the improved carbon based secondary dish support structure (SDSS) for the DVA–2. In DVA–1, the secondary support structure was directly connected to the secondary reflector at four points. At various gravity loads, it is observed in finite element analysis (FEA) that the distortion from the feed platform and adjacent structures are directly transferred into the secondary rim and eventually on to the surface. To separate the effects, a ring made out of carbon composite was placed between the support structure and the secondary reflector. To investigate the size of the ring and especially the layup of the composite, a topology optimization and free-size optimization was performed. A further improvement was achieved by carefully investigating the deformations in the ring and locally stiffening the connection points of the landing tubes on the ring. All these changes in the SDSS resulted in a 96% reduction in RMS residual error for the worst case condition at 15° elevation angle. A combination of careful analyses and application of optimization techniques was paramount to achieve 50GHz performance.
The Next Generation Very Large Array (ngVLA) project to replace the VLA telescope in New Mexico is just beginning. As a part of the initial Community Studies phase, we have contributed the concept design of a 15m feed-low wheel and track design. This telescope, the Next Generation Dish Verification Antenna 15m (ngDVA-15) follows on from the DVA-1 and DVA-2 antennas developed at the Dominion Radio Astrophysical Observatory (DRAO) between 2012 and the present day. This paper will concentrate on the design and optimization process for the ngDVA-15 back-up structure. Topology and free-size optimization were used to develop the initial design concepts. Both methods helped to steer the back-up structure in the initial design phase, but ultimately engineering intuition also played a role. Topology optimization can lead directly to useful solutions in some cases but hardware and software limitations still limit the physical size of the model. Also, topological routines cannot yet correctly model truss-type networks with no moment transfer at the joints, and optimizing structures with only gravitational loads proved to be challenging for the current generation of optimization routines. Size optimization was also used once the design was sufficiently refined. The initial stage of design involved minimization of reflector surface deflections under gravitational loads only. FEA modelling of surface deflections together with in-house developed fitting algorithms were used to determine primary surface accuracy. Surface accuracies of better than 80 microns RMS were achieved which met the initial design goal for telescope operation at 120GHz.
Piezoelectric actuators are a popular choice in micro- and nano-positioning devices. Traditional sensorless position
control approaches use a hysteresis mapping between voltage and position in a voltage feedforward control scheme.
However, this mapping is affected by frequency, temperature, aging etc. Recently, charge control for positioning is also
attracting interest among researchers due to the linear relationship between position and charge. Conversely, a
sophisticated hardware design is required to minimize charge drift. This limits charge based controllers for practical
applications. In this study, a new self-sensing control technique is proposed which requires neither an accurate inverse
mapping nor a sophisticated charge controller. This technique uses a position estimate that is obtained by fusing a
traditional charge based position measurement with a novel capacitance based position measurement. Upon achieving a
reliable position estimate, it is shown that a traditional PI control scheme is sufficient for tracking applications. Different
wave forms having multiple lifts and rates were tested. Error is reduced upto 75% using a self-sensing feedback control
when compared to open loop actuation. These results compare well with traditional self-sensing control techniques found