Description of the Opto- mechanical Design for the PFC

Description of the Opto-

mechanical Design for the PFC

Project name WEAVE

Release Final: Version 1.1

Date: 06 July 2013

Author(s):

Kevin Dee

Owner:

Don Carlos Abrams

Client:

WEAVE Consortium

Document Number:

WEAVE-PRI-017

To maximise the communities' access to information specific to the

project, it is the policy of the project that documentation should be

shared and made freely available to all stakeholders. While fully

exploiting the dissemination of WEAVE information, the Project

Management Team will ensure that the integrity and trust that are

expected between the stakeholders are maintained. Please do not

distribute this document outside the WEAVE Project Team without

the permission of the WEAVE Project Office.

PFC Opto-mechanical Design Date: 06-Jul-2013

WEAVE-PRI-017: Version 1.10 Page 2 of 35

Document History

Document

Location

Printed on Tuesday, 14 October 2014.

The document can be found at :

http://bscw.ing.iac.es/bscw/bscw.cgi/196211

Revision

History

Revision

date

Version Summary of Changes Changes

marked

18/06/2013 0.10 Document created by Kevin Dee Dee

25/06/13 1.00 Document released Abrams

06/07/13 1.1 180 to change to 130 in tables 5 and 6. 65 deg

zenith angle added fig 23 and 24

Dee

Approvals This document requires the following approvals.

Name Title Approval Date Issue Date Version

Gavin Dalton Principal Investigator

Don Carlos Abrams Project Manager

Kevin Dee System Manager

Distribution This document has been distributed to:

Name Title Issue Date Version

Gavin Dalton Principal Investigator 18-06-13 0.10

Tibor Agócs Optical Engineer 18-06-13 0.10

Don Carlos Abrams Project Manager 18-06-13 0.10

Emilie LHome Optical Engineer 18-06-13 0.10

Johan Pragt System Manager 18-06-13 0.10

Kevin Dee System Manager 18-06-13 0.10



TABLE OF CONTENTS

1 INTRODUCTION 5 1.1 Abbreviations 5 1.2 Purpose 5 1.3 Scope 5

1.4 Documents 5 1.4.1 Applicable Documents 5 1.4.2 Referenced Documents 5

2 OVERVIEW 6

3 WHT TRUSSES 6

4 WEAVE RING TO TELESCOPE TOP END 10

5 PFC CENTRAL CAN WITH VANES 12 5.1 Boundary conditions 13

6 LENS 1 CELL AND STRUCTURE TO CENTRAL CAN PRIME FOCUS

INTERFACE 14

7 SUMMARY OF ERROR BUDGET M1 /TO TOP-END UP TO PFC

INTERFACE 17

8 ANALYSIS AND DESCRIPTION OF LENS 1 AND ITS MOUNTING

CELL 17 8.1 Introduction 17

8.2 Original Lens Mount Design 18 8.3 Original Lens Cell 18 8.4 Improved Design for Lens Cell 1 19

8.5 Lens 1 with Constraints Applied to Mounting Pads 20 8.5.1 Results for Lens 1 with pads constrained (no cell included) 21

8.6 Lens 1 and its Cell 23

9 ANALYSIS OF THE EFFECTS OF CHANGING THE SETUP OF THE

RTV PADS 25

10 FEA ANALYSIS OF STRESSESS IN LENS 1 AND ITS CELL 25

11 SUMMARY OF FEA DATA ON LENS 1 AND CELL 26

Figure 1 - Boundary conditions analysis of WHT trusses. ........................................................ 7 Figure 2 - The FARO Vantage Laser Tracker mounted on the telescope cube along with the

laser targets (yellow) for distance measurements. ..................................................................... 8 Figure 3 - With no split collars on the trusses a differential decentre of 95 microns is evident

between the top and bottom ends of the telescope. .................................................................... 9 Figure 4 - Split collars are used to stiffen a truss. ...................................................................... 9

Figure 5 - With split collars fitted to the top trusses, there is a differential decentre of 20

microns between the top and bottom ends of the telescope. .................................................... 10 Figure 6 - Blade spring translation general layout for the WEAVE ring. ................................ 11 Figure 7 - Preliminary FEA of blade spring arrangement indicates a 55µ decentre due to

gravity at the horizon. ............................................................................................................... 11 Figure 8 - Vanes and Central Can. Loaded with PFC, Rotator and Fibre positioner. .............. 12



Figure 9 - At a zenith distance of 65 degrees, the interface between the PFC and the Central

Can exhibits a 100-micron decentre. ........................................................................................ 13 Figure 10 - Decentre with preloaded vanes. ............................................................................. 13 Figure 11 - Lens 1 Cell and extension interfaced to the Central Can. ..................................... 14

Figure 12 - FEA of Lens Cell 1 and its extension tube when at zenith. The three mounting

points are clearly identified as points of minimal displacement. ............................................. 15 Figure 13 - FEA of Lens Cell 1 and its extension tube @ 65 degrees from zenith.................. 15 Figure 14 - FEA of Lens Cell 1 and its extension tube @ 65 degrees from zenith.................. 16 Figure 15 - Lens 1 general dimensions and CoG. .................................................................... 18

Figure 16 - Lens 1 and the original lens cell mounting arrangement ....................................... 18 Figure 17 - Lens surface deformations using the original cell mounting arrangement. ........... 19

Figure 18 - Schematic representations of the lens cell mounting arrangements used in the

DECam (top) and MMT (bottom) designs. .............................................................................. 20 Figure 19 - Lens 1 in the revised lens cell. ............................................................................... 20 Figure 20 - Boundary Conditions for Lens 1 ........................................................................... 21 Figure 21 - Lens L1 surface 1 deformation at zenith was 302.33nm. ...................................... 21

Figure 22- Lens L1 surface 2 deformation at zenith was 249.896nm. ..................................... 22 Figure 23 - Lens L1 surface 1 deformation at zenith was 155.82nm. ...................................... 22 Figure 24 - Lens L1 Surface 2 deformation at zenith was 136.25nm. ..................................... 23 Figure 25 - Overall displacement values for L1 and its cell when the telescope is pointing at

zenith. ....................................................................................................................................... 23 Figure 26 - Lens 1 and its cell at 65 degrees ............................................................................ 24

Figure 27 - The tilt due to the RTV pads at 65 degrees zenith distance (left) and the

exaggerated deformation of the RTV pads (right). .................................................................. 25

Figure 28 - Stress analysis on the lens (left) and cell (right) when at zenith. .......................... 25 Figure 29 - Stress analysis on the lens (left) and cell (right) when at 65 degrees zenith

distance. .................................................................................................................................... 26 Figure 30 - General lens layout with Lens 2 highlighted. ........................................................ 28 Figure 31 - The PFC middle section (for ADC) and end section (for Lens 6) housings ......... 28

Figure 32 - Lens 2 in lens cell. ................................................................................................. 29 Figure 33 - FEA of lens when mounted in its cell at 65 degrees zenith distance. ................... 29

Figure 34 - FEA of lens when mounted in its cell at zenith. .................................................... 30



1 INTRODUCTION

WEAVE is a new wide-field spectroscopy facility proposed for the prime focus of the 4.2m

William Herschel Telescope. The facility comprises a new 2 degree field of view prime focus

corrector with a 1000-multiplex fibre positioner, a small number of individually deployable

integral field units, and a large single integral field unit. The IFUs and the MOS fibres can be

used to feed a dual-beam spectrograph that will provide full coverage of the majority of the

visible spectrum in a single exposure at a spectral resolution of ~5000 or modest wavelength

coverage in both arms at a resolution ~20000. The instrument is expected to be on-sky by

2017 to provide spectroscopic sampling of the fainter end of the GAIA astrometric catalogue,

chemical labelling of stars to V~17, and dedicated follow up of substantial numbers of

sources from the medium deep LOFAR surveys.

1.1 Abbreviations

The abbreviations and acronyms used in this document can be found in WEAVE-MAN-001.

1.2 Purpose

The purpose of this document is to describe the opto-mechanical aspects associated with the

mounting of the WEAVE prime focus corrector (PFC) optics with respect to the telescope

primary mirror (M1).

1.3 Scope

The document presents the results of the analyses of the structures supporting the PFC with

respect to M1. It also looks at the detail mounting arrangement of lens 1 within its cell.

1.4 Documents

1.4.1 Applicable Documents

Document Identifier Document Title

WEAVE-MAN-001 Abbreviations and Definitions

WEAVE-SCI-001 Science Requirements Document

WEAVE-SYS-001 Instrument Development Specification Document

1.4.2 Referenced Documents

Document Identifier Document Title

WEAVE Top End 12006-

900001

WEAVE PROJECT

FEA Services for Top End Ring

Assembly and Focus Can/Centre

Section

Laser tracker Spie 2012-

8444-196.pdf

Using a laser tracker for active alignment on the

Large Binocular Telescope PFC_design_corrector_housing

OCA June 2013.pdf Corrector Housing

WEAVE_report_I.ppt FEM analysis of Lens 1. Report 1

WEAVE_report_II.ppt FEM analysis of Lens 1. Report 2

WEAVE_report_III.ppt FEM analysis of Lens 1. Report 3

WEAVE_report_IV.ppt FEM analysis of Lens 1. Report 4

2009 Blanco DECam design

alignment.pdf

The design and alignment of the DECam lenses

and modelling of the static shear pattern and its

impact on weak lensing measurements

http://www.ing.iac.es/bscw/bscw.cgi/d240354/WEAVE-MAN-001.pdf

http://www.ing.iac.es/bscw/bscw.cgi/d240354/WEAVE-MAN-001.pdf

http://www.ing.iac.es/bscw/bscw.cgi/d252187/WEAVE-SCI-001%20Science%20Requirements.pdf

http://www.ing.iac.es/bscw/bscw.cgi/d260376/Instrument%20Development%20Specification.pdf



CFA Harvard design of cell

mount.pdf

Design of a cell for the wide-field corrector for

the converted MMT

CFA Harvard mounting large

lens MMT.pdf

Mounting large lenses in wide-field instruments

for the converted MMT

13001-900001 FEA for

corrector Lens 1 and cell.pdf

WEAVE PROJECT FEA Services for Corrector

Lens Mounts.

2 OVERVIEW

The optics for the WEAVE prime focus corrector are in the process of a final design review.

The mechanical designs presented for the lens mounts and supporting structures shall

sufficiently demonstrate that the PFC remains within decentration, tilt, and axial spacing

budgets and that induced stresses, and lens surface deformations are tolerable. A level of

design detail is expected which can justify that the designs are understood and although final

opto-mechancal designs are not offered they infer a route to final design that does not

preclude purchasing the optical blanks.

FEA was used to ascertain if the flexure of the WEAVE top-end meets the specifications for

decentre and tilt. If a specification was not met details of the subsequent work required is

described in order to achieve compliance.

Lens 1 is a large lens and more detailed analysis of mounting it in its cell is explored to

demonstrate that the current design is feasible and can ultimately be taken to a final design.

The scope of this report does not include detail descriptions of mechanical systems unless

they have direct influence on procurement of lens blanks.

3 WHT TRUSSES

The performance of the telescope trusses has, to date, been analysed using the Finite Element

Method (FEM). The overall flexure as a function of elevation of the primary mirror (M1)

assembly with respect to the top-end assembly was analysed to derive how much of the error

budget was attributed to the telescope trusses.

The tests were conducted from zenith to the astronomical horizon (90 degrees) but the results

were narrowed to look at a 60 degree range from zenith with 30 degrees zenith distance being

the mean position. The actual observing range requirement for maintaining the optical quality

is zenith to 60 degrees zenith distance (IDS-806).



Figure 1 - Boundary conditions analysis of WHT trusses.

Figure 1 illustrates the boundary conditions that were applied during the FEM analysis. The

telescope is balanced about the altitude axis i.e. as mass and moments are altered at the top-

end of the telescope, the mass is altered at the M1 bottom end to keep the CoG acting through

the altitude rotation axis. Note number highlighted in red are carried forward to table 8

Mean

Angle (degrees). 90 80 70 60 50 40 30

Decentre Lower Truss mm 0.04 0.16 0.29 0.50 0.60 0.78 0.88

Decentre Upper Truss mm 0.01 0.08 0.15 0.22 0.28 0.33 0.37

Net Decentre mm 0.03 0.09 0.15 0.28 0.33 0.45 0.51

FEA Net decentre from mean. Microns 250µ 190µ 130µ 0 -50µ -170µ -230µ

Tilt Lower Truss to Upper truss

-

0.0002 0 0.0003 0.0004 0.0006 0.0007 0.0008

Table 1- Decentre of Telescope Trusses

Top End Displacement/tilt Neg Pos Total Comments

Axial displacement (microns) 10µ 10µ 20µ Focus translation of top end

Decentre x (microns) 25µ 25µ 50µ ALT/AZ not considered an issue

Decentre y (microns) 200µ 200µ 400µ Main component of decentre

Tilt x (degrees) -0.0007 0.0007 0.0014 ALT/AZ not considered an issue

Tilt y (degrees) -0.003 0.003 0.006 Active tilt compensation 0.001 deg

Table 2 - WEAVE optical corrector error budget for M1 to Lens 1 in the PFC. These values were

extracted from the Excel spread sheet (Appendix A).



The FEM of analysis alone is not sufficient to conclude the actual performance of the trusses

and considering that the current performance of M1 to the secondary performs well, it tends to

infer that there is a possibility that the truss FEA analysis does not reflect the actual case. It is

therefore essential that detailed measurements are taken.

The next step is to carry out a detailed survey of the telescope incorporating laser tracking

metrology to see how close the FEA is to reality. Initial investigations suggest that this can be

achieved with the FARO Vantage Laser Tracker placed on the telescope cube (see Figure 2).

Figure 2 - The FARO Vantage Laser Tracker mounted on the telescope cube along with the laser targets

(yellow) for distance measurements.

The FARO tracker was selected because of its ability to operate at different gravity vectors

and to measure with low-level errors (see Table 3).

Table 3 – The FARO Vantage Laser Tracker performance specification.

Laser Tracker

Laser Target



Below is an extract from “Using a laser tracker for active alignment on the Large Binocular

Telescope” (Rakich et al. (2012), SPIE):

“Experience with the tracker at LBT, has shown that the laser tracker is capable of measuring optical

component positions during telescope use with accuracies of the order of 20 microns RMS. Obviously if the laser

tracker could be used to routinely position optics to this degree of accuracy, large positioning errors arising

from large structural gradients would be effectively eliminated, and the problems arising from optics positioning

uncertainties caused by the degeneracy of coma-corrected optics positions would be mostly mitigated”.

Some thought has already been given to improving the decentre values if the actual decentre

of the trusses is greater than the portion of the error budget allocated. However, because the

current M1 to secondary performance is acceptable it is thought that such an improvement

will not be required. Nevertheless, the solution is to fine tune the differential flexure by

stiffening the trusses with a split collar. This has been analysed and the results are promising

as shown in Figures 3 to 5 where the differential decentre is reduced from 95 to 20 microns

Figure 3 - With no split collars on the trusses a differential decentre of 95 microns is evident between the

top and bottom ends of the telescope.

Figure 4 - Split collars are used to stiffen a truss.



Figure 5 - With split collars fitted to the top trusses, there is a differential decentre of 20 microns between

the top and bottom ends of the telescope.

Table 4 shows the error budgets for the performance of the trusses. The budgets have been

calculated to cover an angular range from zenith to 65 degrees zenith distance.

Neg Pos Total Comments

FEA derived decentre values for trusses -230µ +250µ 480µ M1 to telescope top fixed ring

Target decentre values for the trusses -110µ 110µ 220µ

After measurements and optimisation

of truss performance. M1 to telescope

Fixed ring.

Optical error budget allocation M1 to L1 -200µ 200µ 400µ Total error budget allocation M1 to L1

Remaining error budget. M1 to L1 -90µ 90µ 180µ Excluding trusses.

Table 4 - Summary of error budgets for the truss performance.

4 WEAVE RING TO TELESCOPE TOP END

To compensate for thermal expansion and mechanical flexures, the whole WEAVE ring

assembly will translate along the optical axis with respect to M1. In addition to this, a tilt

function will be applied dependant on elevation angle. For the purpose of the optical FDR

only the mounting method flexure as a function of elevation angle is described. The WEAVE

ring translation system, consisting of a series of blade springs (see Figure 6) will be reviewed

separately.

As shown in Figure 7, the total decentre (from the optical axis) with a simulated WEAVE

load (6 tonnes) at 90 degrees zenith distance equates to 55 microns.



Figure 6 - Blade spring translation general layout for the WEAVE ring.

Figure 7 - Preliminary FEA of blade spring arrangement indicates a 55µ decentre due to gravity at the

horizon.

Neg Pos Total Comments FEA derived decentre values for Blade

Springs. -20 +20 40µ

55µ over 90 degrees. Assume 40 microns

over 65 degrees.

Remaining error budget. M1 to L1 -90 90 180 Value excludes Trusses.

Allocated Error budget for Blade

Springs. -25µ +25µ 50µ

Blade Springs Interface. Top fixed ring

to WEAVE ring.

Remaining error budget. M1 to L1 -65µ 65µ 130µ Value excludes Trusses & Blade Springs.

Table 5 - Summary of error budgets with respect to WEAVE top ring (blade springs) over a tolerance

range from zenith to 65 degrees zenith distance



The top ring focus translation resolution of +/- 10 microns is a value required for image

quality.

5 PFC CENTRAL CAN WITH VANES

The next mechanical structure analysed using FEM was the vanes and Central Can up to and

including the PFC mounting interface. The vanes were constrained at the interface to the

WEAVE ring and then each sub system i.e. PFC, Rotator and Fibre Positioner was included in

the model. Each of these systems was simplified to represent their interface, mass and CoG.

From this analysis an error budget allocation was obtained for the WEAVE ring/Vane

interface to the Central Can/PFC interface.

Figure 8 - Vanes and Central Can. Loaded with PFC, Rotator and Fibre positioner.

PFC and

housing

PFC

interface

Rotator

Fibre

Positioner

Central

Can



Figure 9 - At a zenith distance of 65 degrees, the interface between the PFC and the Central Can exhibits

a 100-micron decentre.

5.1 Boundary conditions

There were two principal boundary conditions. Firstly, the vanes were not preloaded as shown

in figure 10, but constrained at the interface between the vanes and the WEAVE ring.

Secondly, the Central Can, vanes and rotator were not optimized to improve the stiffness to

weight ratios.

Figure 10 shows that with the preloaded vanes, the interface between the PFC and the Central

Can undergoes a 70-micron decentre.

Figure 10 - Decentre with preloaded vanes.



The level of preload in each vane was calculated to be 8870N and the stress in the vane was

calculated to be 2.957MPa.

Table 6 shows the decentre values for a 65-degree change in elevation angle. Values are

expressed as +/- values with reference to 32.5 degrees zenith distance.

Neg Pos Total Comments FEA derived decentre values for vanes,

Central Can/PFC interface. -35µ +35µ 70µ 70µ over 65 degrees.

Remaining error budget. M1 to L1 -65µ 65µ 130µ Value excludes Trusses & Blade Springs.

Allocated decentre values for Vanes,

Central Can/PFC interface. -35µ +35µ 70µ

This value includes the Vanes through to

the Central Can/PFC interface.

Remaining error budget. M1 to L1 -30µ 30µ 60µ

Value excludes Trusses, Blade Springs

Vanes and central can.

Table 6 – Decentre of the interface between the PFC and the WEAVE Ring.

6 LENS 1 CELL AND STRUCTURE TO CENTRAL CAN PRIME FOCUS

INTERFACE

Figure 11 - Lens 1 Cell and extension interfaced to the Central Can.

This analysis examines the mechanical decentre and tilt of Lens Cell 1 and the cell extension

which interfaces to the Central Can. The boundary condition constrains the 3 mounting points

to the Central Can interface. This analysis does not include the deformation of Lens 1 (that is

discussed in section 8 of this document), it only considers the general decentre and tilt of the

mechanical structure and its contribution to the overall error budget.

Two of the three mounting

points for Lens 1

Lens 1 extension tube

Lens 1

Central

Can

interface



Figure 12 - FEA of Lens Cell 1 and its extension tube when at zenith. The three mounting points are

clearly identified as points of minimal displacement.

Figure 12 shows Lens Cell 1, with its cell and extension tube interfaced with the PFC/Central

Can section. The PFC/Central Can section is constrained about its 3 mounting points which

gives rise to a translation of 6 microns towards M1 when at zenith.

Figure 13 - FEA of Lens Cell 1 and its extension tube @ 65 degrees from zenith.



As shown in Figure 13, the decentre from the optical axis is 22 microns when the telescope is

at a zenith distance of 65 deg.

Figure 14 - FEA of Lens Cell 1 and its extension tube @ 65 degrees from zenith

As shown in Figure 14, the tilt, with respect to the optical axis, is 0.00093 degrees and the tilt

at 550 mm radius from the optical axis is 9 microns.

Neg Pos Total Comments FEA derived decentre values for lens 1 cell

and extension tube. -11µ +11µ 22µ 22µ Total Decentre from FEA.


Value excludes Trusses, Blade Springs

Vanes and central can.

Allocated decentre values for lens 1 cell and

extension tube. -20µ +20µ 40µ Decentre from Optical axis.


Value Vanes, Central can and structure

support L1.

Table 7 - Decentre of Lens Cell 1 over a tolerance range from zenith to 65 degrees zenith distance



7 SUMMARY OF ERROR BUDGET M1 /TO TOP-END UP TO PFC INTERFACE

This analysis is only preliminary and is intended to demonstrate how each component

contributes to the flexure error budget of maintaining the position of L1 with respect to M1.

The provisional figures are intended to provide a level of confidence that the final error

budget can be achieved for WEAVE. Development and optimization of the top-end structures

are essential as the preliminary designs are taken through to final designs.

Top End Displacement Neg Pos total Comments

Error Budget Allocation M1 to L1 -200µ +200 400µ

Total budget from optical error budget

excel spread sheet.

Target decentre values for the trusses -110µ 110µ 220µ

After measurements and optimisation

of truss performance. M1 to telescope

fixed ring.

Allocated Error budget for Blade Springs. -25µ +25µ 50µ

Blade Springs Interface. Top fixed ring

to WEAVE ring.

Allocated decentre values for vanes, Central

Can/PFC interface. -35µ +35µ 70µ

This value includes the Vanes through

to the Central Can/PFC interface.

Allocated decentre values for lens 1 cell and

extension tube. -20µ +20µ 40µ Decentre from Optical axis.

Remaining Error Budget M1 to L1. -10µ 10µ 20µ

Table 8 - Summary of mechanical flexure error budgets for structures M1 through to L1.

8 ANALYSIS AND DESCRIPTION OF LENS 1 AND ITS MOUNTING CELL

8.1 Introduction

Lens 1 is a large fused silica meniscus lens that needs to be mounted in a lens cell. The

mounting method adopted has to demonstrate that not only does it keep the lens within the

overall decentration and tilt error budgets with respect to M1 but it also needs to show that

induced stresses, and lens surface deformations are tolerable.

The following FEA results show how the original lens mount was not adequate and that the

latest chosen design, which is based on existing designs, is a much more suitable approach.

This analysis makes reference to lens surface deformation which is affected by the mounting

method. Detailed analysis of any lens surface deformation and its effect on image quality is

not discussed in this document but is dealt with in the document entitled “Final Design for the

WEAVE Prime Focus Corrector”.

As Lens 1 is the largest of the six lenses the assumption was made that if a suitable mounting

arrangement (that fulfils the requirements for this lens) could be identified, then it’s likely that

a suitable solution for the other lens mounts could also be identified (see Appendix B).



8.2 Original Lens Mount Design

Figure 15 - Lens 1 general dimensions and CoG.

FEA was performed to model the deformation of Lens 1 in its cell. Different lens cell mount

designs and concepts were analysed until an appropriate configuration was obtained. As this

analysis dealt with the critical glass support interface (of a very large lens) it was deemed

appropriate to contract two independent experts to perform the analysis. The final reports are:

13001-900001 FEA for corrector Lens 1 and cell. -OpTIC Glyndŵr Ltd, Ffordd

William Morgan, St Asaph Business Park, ST ASAPH, LL17 0JD, North Wales, UK

WEAVE_report_IV.ppt FEM Analysis of Lens 1. Konkoly Observatory, Hungary.

8.3 Original Lens Cell

Figure 16 - Lens 1 and the original lens cell mounting arrangement

Fused Silica Lens

231 kg

http://bscw.ing.iac.es/bscw/bscw.cgi/d338381/13001-900001%20FEA%20for%20corrector%20Lens%201%20and%20cell.pdf

http://bscw.ing.iac.es/bscw/bscw.cgi/d338390/WEAVE_report_IV.pdf



Both experts ascertained that the original cell design was inadequate on two counts:

1) Firstly, the three-point mounting arrangement, so close to the lens, deformed the lens

surface to such an extent that the required image quality was not maintained.

2) Secondly, the design did not provide an athermal solution which would maintain

image quality. Furthermore, over large temperature variations the design would

potentially put the optic at risk.

The deformation of the lens surface was 4

microns when held in the original cell

design at zenith.

The deformation of the lens surface was

2.7 microns when held in the original cell

design at 65 degrees from zenith. Figure 17 - Lens surface deformations using the original cell mounting arrangement.

8.4 Improved Design for Lens Cell 1

An improved design based on those used for the MMT and Dark Energy Camera (DECam)

was pursued (see Figure 18).



Figure 18 - Schematic representations of the lens cell mounting arrangements used in the DECam (top)

and MMT (bottom) designs.

Figure 19 - Lens 1 in the revised lens cell.

8.5 Lens 1 with Constraints Applied to Mounting Pads

The boundary constraints were applied as detailed in Figure 20. Fixed constraints were

applied at the locations where the lens cell interfaces to Lens 1. Gravity was applied as a

vector acceleration of 9.81 m/s2 to simulate each of the telescope tube orientations.



Figure 20 - Boundary Conditions for Lens 1

8.5.1 Results for Lens 1 with pads constrained (no cell included)

The results for the Lens 1 analysis are shown in the figures below. The nature of these

deflections is greatly affected by the altitude angle. The aberrations caused by deflection due

to self-weight can be seen to change from power to coma as the zenith angle changes. The

magnitudes are given adjacent to each plot.

Figure 21 - Lens L1 surface 1 deformation at zenith was 302.33nm.

Fixed

Constraints



Figure 22- Lens L1 surface 2 deformation at zenith was 249.896nm.

Figure 23 - Lens L1 surface 1 deformation at zenith angle 65 deg was 155.82nm.



Figure 24 - Lens L1 Surface 2 deformation at zenith angle 65 deg was 136.25nm.

These results illustrate the deformations due to gravity for each of the optical surfaces. The

values range from 137nm to 302nm over the useful range which is specified as zenith angles

of 0o - 65o. Furthermore, there will be residual forces from the support cell that will degrade

the performance of the optic. As such, the values shown in figures 21 to 24 should be

considered as the most optimistic that can be achieved by any mounting arrangement.

8.6 Lens 1 and its Cell

FEA plots were produced for Lens 1 and its cell, when the telescope is pointing at zenith and

at 65 degrees from zenith.

Figure 25 - Overall displacement values for L1 and its cell when the telescope is pointing at zenith.



At zenith, Lens 1 and its cell exhibit a 3-micron displacement towards M1. This is a focus

term. This analysis was ported to Zemax to extract detailed lens surface deformations. The

scope of this document does not include detailed analysis of lens deformation this is covered

in the WEAVE-PRI-014 document. The following is an extract from that document which

summarises the P-V deformations.

“FEA results for lens 1 surface 1 and surface 2, respectively, when the telescope is pointing at

zenith. The difference between the nominal and deformed surface shape is plotted in both

cases. The P-V deformation of the lens is 0.8 and 0.9 micron for the two surfaces”

Figure 26 - Lens 1 and its cell at 65 degrees

At 65 degree zenith distance, Lens 1 and its cell exhibit a 13-micron maximum displacement.

From Table 8, the error budget allocation is +/- 20 microns. This analysis was ported to

Zemax to extract detailed lens surface deformations. The scope of this document does not

include detailed analysis of lens deformation this is covered in the WEAVE-PRI-014

document. The following is an extract from that document which summarises the P-V

deformations.

“For surface 1 and surface 2 the P-V of the difference between the nominal and the deformed

surface shape is 6.5 and 3.4 microns respectively.”



9 ANALYSIS OF THE EFFECTS OF CHANGING THE SETUP OF THE RTV PADS

Further analyses with two sets of radial RTV pads produced good results with respect to lens

deformation but gave a very large tilt of the whole lens due to the RTV shearing which is not

helped by the lens CoG acting about the radial pads. With the telescope at 65 degrees zenith

distance, the displacement induced by the RTV pads was 388 microns.

Figure 27 - The tilt due to the RTV pads at 65 degrees zenith distance (left) and the exaggerated

deformation of the RTV pads (right).

The large scale tilts seen with this design are caused by the centre of gravity of the lens being

located approximately 87mm in front of the radial supports and the compliance of the RTV

material. Additionally, fitting the extra RTV gasket (see figure 29 right) to help with CTE

compensation has had a large detrimental effect on the displacement value. Unfortunately due

to the shape of the lens it is impractical to locate the radial supports at any other location. The

importance of RTV shape and thickness is critical. Calculating and defining the pad sizes for

CTE compensation, minimal displacement and stress is a complex and time consuming

process. This has been achieved before and is well documented in the literature. A first pass

calculation to size the shape and thickness of the RTV pads for Lens 1 and its cell is shown in

Appendix C.

10 FEA ANALYSIS OF STRESSESS IN LENS 1 AND ITS CELL

Stress analysis of the lens cell structure and RTV pads to the lens was conducted with the

telescope at zenith and 65 degrees zenith distance. Figures 28 and 29 show the results of the

stress analysis at the two different angles.

Stress on Lens 0.227 MPa Stress on Cell 6.32 MPa

Figure 28 - Stress analysis on the lens (left) and cell (right) when at zenith.



Stress on Lens 0.19 MPa Stress on Cell 10.11 MPa Figure 29 - Stress analysis on the lens (left) and cell (right) when at 65 degrees zenith distance.

11 SUMMARY OF FEA DATA ON LENS 1 AND CELL

The improved design, using RTV pads and extending the three-point mounting arrangement

away from the lens, helps to reduce the lens surface deformations to an acceptable level.

Overall, displacement of the lens and lens cell fall below the error budget allocations given in

Table 8. Tilt and focus of L1, with respect to M1, is provided by actuation of the new

WEAVE top-ring.

FEA data collected when using RTV pads and RTV gaskets is poor with respect to tilt. The

importance of the “shape factor” (load surface area to edge surface area) is not only critical

for an athermal design but also for overall displacement of the lens in its cell.



Appendix A - Flexure Error Budget taken from the error budget spreadsheet.

Comp Unit

NEG

TOL

POS

TOL

Flexures 2 - Lens displacement due to gravity load - axial displacement can be compensated by TOP-END refocusing, tilt by optimizing to degrees zenith angle

LENS 1

lens displacement/tilt

axial displacement µm -25 25

decentre x µm -10 10

decentre y µm -70 70

tilt x deg -0.001 0.001

tilt y deg -0.01 0.01

ADC 1





tilt x deg -0.001 0.001

tilt y deg -0.01 0.01

ADC 2





tilt x deg -0.001 0.001

tilt y deg -0.01 0.01

LENS 6





tilt x deg -0.001 0.001

tilt y deg -0.01 0.01

IMA

IMA plane displacement/tilt


tilt x deg -0.001 0.001

tilt y deg -0.01 0.01

WHT wrt Corrector

top-end displacement/tilt




tilt x deg -7E-04 0.0007

tilt y deg -0.003 0.003


Appendix B – The initial FEA of Lens 2 deformation and the middle and end structures of

the PFC.

Figure 30 - General lens layout with Lens 2 highlighted.

Figure 31 - The PFC middle section (for ADC) and end section (for Lens 6) housings

Lens 2

Middle

Section

End

Section


Figure 32 - Lens 2 in lens cell.

Lens 2 (NBK7) is mounted in the same fashion as Lens 1 using multiple radial and axial pads.

This arrangement incorporates Delrin pads and a steel cell.

material

Diameter

(difference)

Radius

(difference) CTE

refere

nce

temp.

operat

ing

temp.

Radius

shrinkage

mm mm *10^-6 K K micron

LENS 2 axial

lens NBK7 150.000 75 7.1 298 268 16

spacer Delrin 6.500 3.25 120 298 268 12

mount Steel 156.500 78.25 12 298 268 28

total 0 The thickness of the CTE-compensating Delrin radial pads is 6.5 mm.

As shown in Figure 33, a preliminary check of the surface deformation of Lens 2, when the

lens is mounted at 65 degrees zenith distance, shows a deformation of 100nm (P-V) for

surface 1.

Figure 33 - FEA of lens when mounted in its cell at 65 degrees zenith distance.

Similarly, as shown in Figure 34, when the lens is mounted at zenith the deformation of

surface 1 is increased to 200nm (P-V) for the same surface.


Figure 34 - FEA of lens when mounted in its cell at zenith.

Analysis of middle section

Middle section loaded with ADC and cells at zenith

Decentre from optical axis 1.2 microns.

Axial displacement towards “L1” is 14 microns.

Tilt is less than 0.0001 degrees.


Middle section loaded with ADC optics and cells at 65 degrees from zenith.

Decentre from optical axis is 14 microns.

Axial displacement towards “M1” is 6 microns.

Tilt less than 0.0001 degrees.

Analysis of end section

End section loaded, at zenith, with Lens 6 and its cell.

Decentre from optical axis is 1 micron.

Axial displacement towards “ADC optics” is 3 microns.

Tilt is less than 0.0001 deg.


End section loaded with Lens 6 and its cell at 65 degrees from zenith

Decentre from optical axis is 2.3 microns.

Axial displacement towards “ADC optics” is 1 micron.

Tilt is less than 0.0001 degrees.

In summary, the deformation of Lens 2 appears to be acceptable and the flexure error budgets

for the housing are within the error budget allocation detailed in appendix A.


Appendix - C Sizing RTV Pads

