EPIC Medium-Scale Optical Design Huan Tran Brad Johnson Mark Dragovan April 2009.

EPIC Medium-Scale Optical Design

Huan TranBrad Johnson

Mark DragovanApril 2009

EPIC-IM optical layout

EPIC-IM optical properties

•Crossed Dragone•ABS, Clover, QUIXOTE, QUIET ….

•Unprecedented Large FOV• 30x20 degrees

•Extreme Compact design•Maximize resolution/throughput in shroud

•Telecentric•NO refractive elements

EPIC-IM optical properties

“force” telecentric focal plane => cold aperture

Oversize mirrors

EPIC-IM aberration performance

• Elliptical focal plane– Limited by vignetting in Y– Limited by aberration in X

• Multiband– High Frequencies in center– 30—800 GHz

• 11,000 bolometers

30 deg/ 160 cm150 GHz

30 GHz

30 GHz

850 GHz

EPIC-IM cold vs warm

Main Beam Simulations

• Beam shapes– No Refracting elements– Calculated with Perfect Gaussian Feed horns– Calculated for each “Hex”

• Polarized beam-scale distortions– Fit Gaussians to beams– Compare to benchmarks

PO co and cross–polbeams for single feed

PO vs GO sanity check30 GHz beam, from Grasp 9

Spot diagram, from Zemax

Main beam effects vs benchmark

EPIC-IM mirrors alone are below benchmarks w/o modulation

Sidelobes

• Analyzed with Physical Optics(PO) an Geometric Theory of Diffraction (GTD)

• Aperture Integration Method– Optics box

• Galaxy Convolution

EPIC-IM straylight

EPIC-IM farsidelobes

Co-pol beam, no baffling

~15d simulation time

Calculate Equiv J

Aperture Integration methodSet J =0 outside

28d sim time

Polarized Far Sidelobes

(QT2 + UT2 + VT2)1/2

2 x15d sim time

Galactic Contamination

• In order to evaluate the effect of the signal from the far sidelobes, we convolve the qt beam maps with a 150 GHz sky model.

• The beam patterns have the primary beam masked, so only the response to the sidelobes are evident in the output.

• The sky data is an all sky map at long wavelengths (150 GHz). • Since the beam is asymmetric, it is necessary to rotate the beam with respect to the sky at

each point to get the complete convolution.• The convolutions were done using the totalconvolver code developed by the Planck

community. (Gorski et. al.)

M.Dragovan

(above) The 150 GHz sky map with which the beams are convolved. Units are log(uK).

(below) The results of the convolution, qt beam with the above sky . Units are (uK).

M.Dragovan

In order to further quantify this result, we make two histograms: the number of pixels with a given intensity (left plot), and the integrated histogram giving the total number of pixels less than a given intensity (right plot). This is similar to the plots that are shown for site surveys.

By inspection one can see that fully 90% of the pixels are <0.2nK.

M.Dragovan

Goal 1nK

GTD and Polarization of sidelobes• Our Far sidelobe simulations were for 3.25 –f –, but

Conclusions

• EPIC-IM has enormous throughput• Systematic beam effects are below benchmark• Sidelobes are manageable

– More analysis time required to be sure