Transcript
Summary of the conceptual phase on “Mirror Cover Repair” P. Salinari, January 19 2009
Reference Documents: Ref. 1: MC_Specs_rev_b_570s002b Ref. 2: LBT Primary Mirror Cover FEA of as built configuration (727a130d) Ref. 3: Mirror Cover problems –rev_ADS.pdf
Foreword The present document simply summarizes the conclusions drawn by INAF‐Arcetri (P. Salinari and L. Miglietta) with the support of ADS International (D. Gallieni, E. Anaclerio, P. Lazzarini) on how to solve the mechanical problems affecting the Mirror Covers of LBT after a first conceptual phase. The work done until now addressed ONLY the mechanical problems of the MC and consisted in:
Collecting the (sparse and sometimes scarcely documented) information on the malfunctioning of the MC (reported in Ref. 3)
Setting up a finite element model of the MC based on “as built” drawings and able to closely reproduce the known measurements (deflections, resonant frequencies) (Ref. 2)
Studying, by means of FEA, various options to resolve the identified MC problems. (Ref. 2)
Identifying a set of modifications to the existing hardware which could resolve the identified problems (Ref. 3 for a full list of considered corrective measures, this document for the proposed ones)
The purpose of the current document is that of summarizing an initial proposal for a discussion with LBTO on the modifications to be done to the existing MC units, so that detailed design could start.
General outline of the Problems and of the proposed solutions The known mechanical problems and a variety of possible solutions are described in Ref.3; here we will summarize the main problems and the proposed solutions taking into account the results of the latest version of the Finite Element Study, Ref. 2 1. The stiffness of the rotating arm is too low. The lack of rigidity is clearly a consequence of the limitations in space (which include the need of anchoring the MC to the telescope in an unfavorable way) and it is worsened by the fact that the “stow pin”, currently located near the rotation axis, can only provide “on the paper” a modest
effect on the system rigidity (and in practice has essentially no effect at all). This is at the origin of many of the MC problems: o when parked the low resonant frequency of the MC affects the telescope drives o during deployment the bending due to gravity originates interference of the free end of the MC with the cell (the remedy of lifting the entire unit by adding shims originatesor just increased the interference between the MC arm and the M3 arm) o when deployed and brought in Horizon pointing position the overturning moment of the cover is almost only balanced by the retaining wheels of the cover (which are not suitable for this heavy load and were damaged) The very limited available space and the limitations in attachment/rotation geometry of the entire MC unit discourage actions oriented to reduce the bending by stiffening the MC components. We therefore looked for solutions where we could use the available space to install additional supports, more effective than the existing stow pin in reducing the moments exerted by gravity on the rotating arm (or, equivalently, reducing the free length of the beam, therefore increasing its resonant frequencies). The FEA work has identified the location of two separate clamps (respectively for the rest position and for the deployed position) which could replace the current stow pin and increase the stiffness of the arm , when clamped in these two positions, by a large factor. A rail supporting the arm during its motion between the two clamps would significantly reduce the deflection problem also during deployment, therefore removing the need of displacing (“shimming”) the unit to avoid interferences with the cell edge. The design leaves a 35 mm gap between an un‐deformed MC arm, while the measured deflection is 32 mm. The rail therefore only needs to reduce the maximum deflection by about a factor of two. These measures (two clamps and a rail) seem to be sufficient to remove or to greatly alleviate problems 0, 2, 3, 4, 6 reported in Ref 3. 2. The cover “blades” interfere with the M3 support. Even if the other interferences originated by the excessive flexure of the MC arm can be removed by the measures mentioned above, this problem forces in any case to change the blades of the cover. This change of the blades, in turn, helps in gaining larger margins for solving the problems listed in the first group. In particular, thanks to the relaxation of some of the current specifications with respect to the original ones, we can consider not only reducing the height of the blades, but also their total mass (e.g. changing the material from steel to Aluminum alloy or to CFRP composites). The reduced mass of the blades also allows adopting simpler (and more reliable) wheels and clamps for the deployment of the blades.
Proposed corrective measures Here we list the sub‐set of conceptual corrective measures we have identified as sufficient among those considered in Ref.3. Presently there are no drawings at any level of detail for any of the following corrective measures, only the verification that the necessary physical space is available on the telescope and the results of FEA, when relevant, to evaluate the effectiveness of the measures. The code in
brackets is a reference to the description provided in Ref 3 (for instance: “PS 1a” means “Possible Solution 1a” of Ref3). 1. Clamp at rest position (PS 2). (See pg 27 and followings of Ref 2 for a schematic figure and for dynamic analysis results) Although the first resonant mode is at about 9 Hz (vs 10 Hz required in Ref.1) this mode is essentially orthogonal to the excitation provided by the drives and should have no effect. In any case, its frequency can be increased (e.g. by providing a larger angle between the two beams in the “inverted V” geometry of Ref.2). The first mode affecting the telescope drives is already above 10 Hz (10.4) in the FEA model and, taking into account the possibility of reducing the MC mass (see measure 4), it seems that the 10 Hz requirement of the specifications is achievable in practice. In the subsequent design phase, the clamp will need to be optimized to increase first of all the resonant frequency of modes coupled with the drives, i.e. of modes with a large component of motion parallel the Z axis. The type of clamp must be such to restrain all motion components (including rotations). This requires a loading force and an appropriate shape of the contact (e.g. sphere‐cone). 2. Clamp at deployed position (PS 4). (See 727a130c pg30 and followings) The clamp shown in Fig 30 of Ref. 2 is sufficient to reach the specs in terms of frequency in deployed position. The above comment on the effect of a reduction of the mass applies to the present case too. Even in absence of static analysis, the total deflection at the free end of the arm can be evaluated to be of the order of 1 cm. Also this clamp must be able to restraint all degrees of freedom and requires a loading force and an appropriate shape for the surfaces in contact. 3. Rail to support the arm during rotation (PS 3c). The basic idea is that of placing a rail (shaped as an arc of a circle of about 70°, with center on the arm rotation axis) on the telescope wind‐bracing supporting the entire MC unit and, on the rail, a wheel (placed approximately under the intersection of the long and short arms of the T beam), so that the rail supports a significant fraction of the weight of the arm during the rotation, essentially removing the bending moment responsible of the flexure of the rotation mechanism (see fig 5 of Ref. 2) and reducing also the flexure of the T arm, whose “flexure length” is significantly shortened . The function of this support is that of reducing and controlling the large deflections of the arm during the motion between the clamps 1 and 2 that are currently present.
By adjusting appropriately the height, inclination (and, if necessary, out of plane curvature) of the rail it is in principle possible to control the arm trajectory above the cell edge during its rotation and the correct engaging of the clamps at either end of the motion. 4. Replacement of the cover blades. (PS 1a) The new blades will be of a trapezoidal shape similar to the existing ones, but of reduced height (at least by 30 mm) and therefore in a larger number to preserve a reasonable angle (say about 90°) when fully open. Due to the reduced requirements on the loads the cover needs to support, the new blades must be designed only to resist an impact as described in the specifications. Thermal requirements (i.e. avoiding strong vertical thermal gradients in the mirror) are satisfied by a single sheet of even a highly thermally conductive material (such as Aluminum Alloy or CFRP) and water and dust tightness can be satisfied by a strip of impermeable and flexible material (plastic sheet, canvas) glued to the faces and enclosing the hinges. So there is no need of re‐building the existing canvas cover. Other aspects of the cover can be simplified by replacing the blades:
The wheels rolling on the cell flange can be integrated in the lower hinges. This provides a more reliable attachment, allows for smaller wheel size and this, in turn, helps in reducing the space needed when the blades are retracted.
The retention wheels rolling on the lower edge of the cell flange can be replaced by “hooks” of appropriate “slanted” shape (so to be unaffected by the residual welding present on that face of the flange) which could be built as an integral part of one half of the blades.
5. Other minor changes (PS 0a) (PS 3a). These changes will only be applied If necessary, e.g. to gain extra margins in clearance:
the pin of the M3 support can be installed upside‐down (PS 0a)
the reduction of the vertical extension of the central pin of the MC (PS 3a)
Expected compliance with the Specifications Although performances will be better evaluated at a more advanced stage of the design, the proposed changes already show the potential for satisfying the basic mechanical requirements reported in the specifications. Here we briefly discuss some of them:
Spec 8.1. Protect Mirror from Impact Damage. The request of withstanding at least the impact of 1 kg falling from 5 m height should be satisfied easily by a 1 mm thick Al blade
riveted between two steel hinges, but a test is required before taking a decision on the material and its thickness. Spec 9.1. Mass. The current total mass of about 5000 kg for both units is likely to be preserved or reduced. The change of the deployable covers from steel to a lighter material of equal thickness and the simplification of the wheels and retention mechanisms will reduce the total mass by more than 1000 kg, while the added clamps and supports should not reach this value. Spec 9.2. Resonant Frequency. As discussed above the specified resonant frequencies seem to be achievable.
It is worth saying explicitly that we still have doubts that even with the proposed changes the MC will be reliably deployable at telescope positions far from Zenith Pointing (defined as “desirable” in Ref. 1). Although the proposed mechanical changes are in a direction which favors deployment and stowage at lower elevation angles, they will likely be not sufficient and moreover we haven’t examined the problems connected with the required torques (therefore motors) and load capability of the rotation mechanics.
Further steps While the obvious next step is reaching a consensus on a baseline for further and more detailed design work, other actions are urgently needed, in particular:
Quantitative verification of the assumptions (some doubts remain on the measurements of interferences and on deflections) Verification of the general conditions of the MC mechanical parts which we propose to use as they are (the MC have been stored at the Base Camp for years in open air, protected by a canvas cover) Some (less urgent, but necessary) actions are needed on LBTO side to verify aspects which have not been considered until now, in particular concerning the functionality and compliance with the specs of the existing electrical and control systems of the MC. These “electrical” subsystems will not be part of the work planned by INAF and LBTO needs to identify and perform the actions deemed necessary.
Conclusions Although with some remaining uncertainties, the conceptual phase summarized here has shown that the identified mechanical problems of the MC could in principle be resolved by using the original design and, largely, the existing mechanical hardware with the addition of an auxiliary support structure (composed by a rail and two clamps) and by replacing the deployable cover with a lighter and simpler one. Therefore, if the verifications mentioned above confirm the assumptions of our conceptual work, there is a perspective for an affordable solution of the MC problem.
