JAXA / GSFC - NASA · 2016-12-23 · ASTRO-H INSTRUMENT CALIBRATION REPORT Mirror Definition File ASTH-HXT-CALDB-MIRROR Version 0.2 December 22, 2016 JAXA / GSFC Prepared by: A. Furuzawa,

ASTRO-H

INSTRUMENT CALIBRATION REPORT Mirror Definition File

ASTH-HXT-CALDB-MIRROR

Version 0.2

December 22, 2016

JAXA / GSFC

Prepared by: A. Furuzawa, H. Awaki, S. Yamauchi, H. Matsumoto, I. Mitsuishi et al., GSFC Hitomi Data Science Center

2-22

.

Table of Contents

1 Introduction ............................................................................................................................... 4

1.1 Purpose .............................................................................................................................. 4 1.2 Scientific Impact ................................................................................................................ 4

2 Release CALDB 20160310 ....................................................................................................... 7 2.1 Data Description ................................................................................................................ 8 2.2 Data Analysis ..................................................................................................................... 9 2.3 Results .............................................................................................................................. 12 2.4 Final remarks ................................................................................................................... 21

3 Release CALDB 20161222 ..................................................................................................... 21 3.1 Data description ............................................................................................................... 21 3.2 Data analysis .................................................................................................................... 22 3.3 Final remarks ................................................................................................................... 22

3-22

.

CHANGE RECORD PAGE (1 of 1) DOCUMENT TITLE : Mirror Definition File

ISSUE DATE PAGES AFFECTED DESCRIPTION

Version 0.1 10 Mar 2016 All First Release

Version 0.2 22 Dec 2016 All Second Release

4-22

.

1 Introduction 1.1 Purpose This document describes the elements and structure of the HXTs related to the telescope performance, implementation of them to the raytracing code (xrtraytrace) and how to create the CALDB files used by xrtraytrace. The CALDB file structure is define in the ASTH-SCT-04 and available from the CALDB web page at http:// hitomi.gsfc.nasa.gov. 1.2 Scientific Impact

As shown in Fig. 1, the HXT consists of the primary and secondary parts and each part is divided into 3 segments containing 213 reflector shells confocally nested. Each of top and bottom edges of reflectors in a segment are positioned by 8 alignment bars arranged radially. A fan-shaped region sectioned by neighboring alignment bars is defined as 'sector'. The radial position of alignment bars were tuned to focus X-rays as narrowly as possible. Each reflector has figure errors in both radial and circumferencial directions. Those figure errors of reflectors and remaining misalignments cause a scattering and blurring images focused by each of local parts of aperture of the telescope. Because the reflectivity strongly depends on the incident angle, we divided the reflectors into 14 groups by their incident angles, and optimized the multilayer design for each group. The reflectivity of the multilayer depend on the multilayer design and interfacial roughness.

Fig. 1 Structure of the HXT

2. Design Parameters

A. Overview

Themission requirements for the ASTRO-HHXTarea total effective area of 300 cm2 at 30 keV, and a HPDof 1.70 at 30 keV [10]. ASTRO-HHXTwill have a larg-er effective area than NuSTAR, while the angularresolution of the HXT will not be better. To meet therequirement for effective area, the ASTRO-H HXTmirrors employ tightly nested, conically approxi-mated thin-foil Wolter-I optics, and the mirror surfa-ces are coated with Pt/C depth-graded multilayerssimilar to those of the InFOCμS telescopes. The basicdesign parameters of the proposed ASTRO-H HXTare as follows: 12 m focal length, 45 cm diameter tele-scope consisting of 20 cm long foils in two (primaryand secondary) stages (see Table 1). The diametersof the innermost and the outermost reflectors are120 and 450 mm, respectively.

The image blurring of thin-foil Wolter-I optics iscaused by the figure errors of individual reflectors,uncertainties of positions of individual reflectors,the image asymmetry caused by an error of the align-ment bar position, and a conical approximation of theWolter-I optics [11]. To achieve the requirement forthe HPD, we assigned the error budget on the HXTHPD as listed in Table 2 from the experience of theballoon-borne experiment.

B. Telescope Housing Design

Detailed design of HXT has been carried out based onthe InFOCμS and SUMIT balloon-borne experiments[11], under the constraint of the space within the

nose fairing of the HII-A rocket. Figure 2 shows aschematic view of the design of HXT. The diameterand height of the HXT housing are 480 and509.4 mm, respectively. HXT has three mount tabsto be mounted on the FOB top plate, and consistsof four components: two TSs, a PC, an x-ray mirror,and a reference cube.

The ASTRO-H HXT mirrors are installed into atelescope housing, and are positioned by alignmentbars as shown in Fig. 2. Each alignment bar has aseries of grooves to hold the mirror reflectors. Whilethe x-ray mirror housing of the Suzaku XRT is sep-arated into primary and secondary parts, both partsof the mirror housing are unified in the ASTRO-HHXT. As a result, the bottom edge of the primary re-flectors and the top edge of the secondary reflectorsare held at their desired locations by the same align-ment bars. Thanks to the unification, we can avoidmisalignment between the primary and secondaryreflectors. The alignment error is expected to be afewmicrometers, which is determined by both manu-facturing error and the groove width of the align-ment bars.

The temperature of the HXT housing and mirrorreflectors will be controlled at 22! 2°C during obser-vations using a heater power control system. Thetemperature of 22°C is the room temperature ofthe clean room at Nagoya University, where the HXTwas fabricated. To provide temperature stability ofthe HXT and to reduce temperature control heaterpower in orbit, the HXT housing is designed to bethermally isolated. We use a thermal insulator madeof titanium alloy to reduce the thermal conductivitybetween the HXT housing and the top plate of FOB,and both the top and bottom apertures of the tele-scope are covered with TSs. Furthermore, to makethe temperature of each mirror reflector uniform,the inner surface of the housing is black anodized.Details of the thermal design of HXT are describedby Ito et al. [12].

After measuring the relative orientation of theHXT optical axis and the reference cube in a ground

Table 1. Design Parameters of ASTRO-H Hard X-RayTelescope

Items HXT

Focal length 12 mNumber of modules 2SubstrateMaterial AluminumSubstrate thickness 200 μmAxial length 200 mmReflectorsMaterial Pt/C multilayerAdhesive material EpoxyAdhesive thickness 20 μmNumber of nesting shells 213Diameter of innermost reflector 120 mmDiameter of outermost reflector 450 mmIncident angle 0.07°–0.27°Number of reflectors/telescope 1278Geometrical area/telescope 968 cm2

Weight/telescope ∼62 kg

Table 2. Error Budget on the HXT HPD

Reflector figure error 1.40

Positional error of reflector 0.70

Bar positional error 0.40

Conical approximation error 0.30

Fig. 2. Schematic view of the ASTRO-H HXT. The right part dis-plays a cross section of HXT.

7666 APPLIED OPTICS / Vol. 53, No. 32 / 10 November 2014

Segment Alignment bar Sector

5-22

.

Thus the telescope performance such as an effective area, a vignetting function, a Point Spread Function (PSF) and an Encircled Energy Function (EEF) strongly depends on those local properties of reflectors. In the raytracing code (xrtraytrace), each reflector is defined as a series of subfoils arranged in a circumference direction, and each subfoil has a different position, orientation, reflection profile and multilayer structure as local properties. The implementation of subfoils and their properties enable us to estimate actual performance of the telescope.

Fig. 2 implementation of subfoil

The CALDB files are used to implement local properties derived from the measurements to reproduce the effective area and the PSF of the HXTs. Contents of CALDB files 1) Mirror file

The mirror file contains following 6 extensions.

i) MIRROR extension : definitions of subfoils and their parameters for reflectors A subfoil is defined as a ‘sector’ part of a reflector defined by alignment bars. The definition of subfoil unit is commonly used for a primary reflector, a secondary reflector and a precollimetor blade. Following parameters are defined for each subfoil. • Positions of edge points in cylindrical coordinate

Design values are stored.

• Names of reflectivity tables for the front side and back side of reflectors A reflectivity table for a front side is calculated for each multilayer group. For a back side, it is defined to refer the common reflectivity table.

systilt

systwistsubfoil

position of revolution axis for systilt is adjustable

z

x

Y

reflector

−2×10−3 0 2×10−30.01

0.1

110

100

1000

104

Prob

abilit

y [/r

ad.]

Scattering angle [rad.]

SP8Seg1_sm1−2_sec5 30(Black,40(R),50(G),60(B)keVReflection Profile

（colors express energies）

Reflectivity table（depend on multilayer and interfacial roughness）

6-22

.

• Name of the reflection profile table for the front side of reflectors

Aperture is divided into regions defined with the boundaries of multilayer groups and sector boundaries (the alignment bars), and the common reflection profile is defined for primary and secondary subfoils in each region.

• Misalignment in the direction of tilt and twist In the current version, they were derived from flux weighted mean of centroids of spot images (see 2.2). They may be tuned according to in-flight EEF (PSF) calibration.

ii) OBSTRUCT extension : definitions of support structures obstructing a photon

Support structure such as alignment bars and sector covers are implemented as a plane blocking photons. The plane is defined by X-Y coordinates of its vertexes and plane’s Z coordinate. Therefore, the plane is parallel to an X-Y plane. Design values are stored.

iii) SEGMENT extension : definitions of misalignment of each segment

Displacement and orientation of each segment from the design value are defined. In the current version, design values are stored (displacement and orientation = 0). They may be tuned according to in-flight effective area, vignetting and EEF (PSF) calibration.

iv) COLLIMATOR extension : definitions of subfoils and their parameters for precollimator blades

Definition of subfoil is the same as that of a reflector. For both front side and back side, the same reflectivity table and reflection profile are defined to be referred as those for back side of reflectors. They may be tuned according to in-flight effective area, vignetting and EEF (PSF) calibration.

v) SURFACE extension : definitions of multilayer structure and its interfacial roughness

Design values of multilayer structures are stored. Roughness parameters are derived to fit the energy dependence of the effective area of each multilayer group according to the ground calibration data. Roughness may be tuned according to in-flight effective area calibration.

vi) AZIMUTHALSTRUCT extension : definitions of thermal shield geometry and materials

Design values are stored. 2) Scatter file: reflection profiles for front side and back side of reflectors and precollimator

blades Front side reflection profiles were determine by on-ground measurement data (see 2.2). They may be tuned according to in-flight effective area, vignetting and EEF (PSF) calibration.

7-22

.

For precollimator blades, a common reflection profile is used for both front and back side. In the current version, below 12 keV, the profiles are the same as those applied for back side reflection of the Suzaku XRT reflectors. Above 12 keV, the profile at 12 keV is applied. The reflection profile for precollimator blades and back side of reflectors will be updated after finishing analysis of the on-ground measurement in future.

3) Reftrans file

It contains reflectivity and transmissivity tables for all multilayer groups calculated by xrtreftable with the current CALDB files, reflectivity tables for back side of reflectors and both side of precollimator blades, and mass absorption coefficient table for the thermal shield and the central cover. For reflection at back side of reflectors and precollimator blades, a common reflectivity table is used. The current version includes the reflectivity table measured with Suzaku XRT reflectors below 20 keV and calculated table assuming aluminum smooth surface above 20 keV. The reflectivity tables for precollimator blades and back side of reflectors will be updated after finishing analysis of the on-ground measurement in future.

4) atomicScattering file

AtomicScattering file provides various tables of atomic scattering factor (f1 and f2) from different literatures. Simulation results in this document and deriving optimum interfacial roughness were obtained with atomic scattering factors in the extension ‘HenkeSskChant’ in the current atomicScattering file. The scattering factor in ‘HenkeSskChant’ come from S.Sasaki 1989 (KEK Report, 88-14, 1-136) and S.Sasaki 1990 (KEK Report, 90-16, 1-143) at energies above 4275.34 eV, and Henke 1993 (Atomic Data and Nuclear Data Tables Vol. 54 (no.2), 181-342) at energies less than and equal to 4275.34 eV. They may be tuned with in-flight data, if it is nessesary.

2 Release CALDB 20160310

Filename Valid data Release data CALDB Vrs

Comments

ah_hx1_mirror_20140101v001.fits 2014-01-01 2016-03-10 001 Mirror ah_hx2_mirror_20140101v001.fits 2014-01-01 2016-03-10 001 Mirror ah_hx1_scatter_20140101v001.fits 2014-01-01 2016-03-10 001 Scatter ah_hx2_scatter_20140101v001.fits 2014-01-01 2016-03-10 001 Scatter ah_hx1_reftrans_20140101v001.fits 2014-01-01 2016-03-10 001 Reftrans ah_hx2_reftrans_20140101v001.fits 2014-01-01 2016-03-10 001 Reftrans ah_gen_atmsca_20140101v001.fits 2014-01-01 2016-03-10 001 atomicScatter

8-22

.

2.1 Data Description Ground calibration data used for deriving parameters stored in CALDB files have been taken at the SPring-8 BL-20B2 in Nov. 2012 for HXT-1 and Oct. 2013 for HXT-2. Details of the ground calibration are described in Awaki et al. 2014 (Applied Opics, 53, 32, 7664). An X-ray beam was collimated into 10 mm x 10 mm rectangular shape and one segment of the HXT was fully covered by a mosaic mapping method (spot scan measurement) with that X-ray beam. Each X-ray beam forms a image (local spot image) which has different centroid and a profile on the focal plane (Fig. 3). It is due to figure error and misalignment of subfoils.

Fig. 3 Spot scan measurement

As described in 2.2, an offset from an ideal focus point was converted into the misalignment parameters, systilt and systwist. Reflection profiles for fromt side of subfoils were also derived from the one dimensional profiles of local spot images. Measurements with the mosaic exposure method have been done at 20 (HXT1 only), 30, 40, 60 and 70 keV. For assessment of atomic scattering factors of Pt and C around Pt L and K absorption edges, reflectivity of Pt/C multilayers has been measured at SPring-8 BL-01B in 2014 (Fig. 4).

66 第 6章 X線を用いた HXT-2の性能評価

-200

-150

-100

-50

0

50

100

150

200

-200 -150 -100 -50 0 50 100 150 200scan_path

Fig. 6.1 ラスタースキャンのスキャンパス。横軸は Y方向、縦軸は Z方向のステージ移動量を表し、望遠鏡中心を原点としたときの位置を mm 単位で表示した。図中に示した正方形の枠が 10mm 四方のビームを照射する領域を表し、この枠内に示した数字がスキャンパスのシーケンス番号を表す。この図は、Fig. 5.8で定義したホールサイド配置におけるビームのスキャンパスを示す。X線は、紙面の奥から手前に向かって入射する。

プルステージを移動させる。このとき望遠鏡と検出器の位置関係を保つために、サンプルステージと同期させて検出器ステージも移動させる。

この測定方法をラスタースキャン法、あるいは単にラスタースキャンと呼ぶ。Fig. 6.1に、ラスタースキャンにおけるスキャンパスの例を示した。なお、望遠鏡を取り付けているサンプルステージや検出器ステージの稼動域に制限があり、一挙に全セグメントのスキャンができないため、HXT-2の性能測定は一つのセグメントごとに行った。測定するセグメントは常に、Fig. 5.8で定義したホールサイドに配置した。

6.2 データの補正 67

Fig. 6.2 測定の概念図。ラスタースキャン法を用いて、一つのセグメントごとに測定した。左図のような 3カ所に X線ビームを当てると、中央図のような細長い反射像がそれぞれ得られる。最終的に全てのパスをスキャンすると、右図のような蝶々型の反射像が出来上がる。

ラスタースキャンによる測定の概念図を Fig. 6.2に示した。ラスタースキャンの各スポット（10 mm四方の X線ビームの当たるそれぞれの領域）に X線が当たると、3.4.3項で議論した理由により焦点面では細長い反射像をつくる。最終的には、全てのスポットで得られた反射像を足し算することで、望遠鏡全体に平行な X線を当てて得られる焦点面像を擬似的に再現できる。

6.2 データの補正測定で取得したある生データ（反射像イメージ）は一般的に、式 (6.1)に示すような成分を含んでいる。

I ′(y, z) = c(y, z)× I(y, z) + dark + bias (6.1)

ここで、(y, z)は検出器面上での座標、I ′(y, z)は取得した反射像イメージのカウント値、I(y, z)は真のカウント値、c(y, z)は検出器がピクセルごとにもつ検出効率、darkは暗電流によるノイズ成分、biasは測定値のもつバイアス成分を表す。求めたい真のカウント I(y, z)を取り出すには検出器の検出効率を一様に揃え、ノイズ成分やバイアス成分を差し引かなければならない。性能測定においては dark（ノイズ成分）と bias（バイアス成分）を区別しないので、まとめてダーク成分と呼ぶ。HXT-2の性能測定はラスタースキャンで行うため、最終的に得られる蝶々型の反射像イメージには式 (6.2)に示す

ような成分が含まれる。

I ′add(y, z, t) =522!

i=1

(fi(t)× I ′i(y, z)) (6.2)

ここで、fi(t)は X線ビーム強度の時間変動を表す。ラスタースキャンでの測定には長時間を要するため、ビーム強度の長期的な時間変化の効果が影響する。つまり真のカウント値 I(y, z)を取り出すには、ダーク成分と検出器の感度ムラ補正に加えて、ビーム強度の長期トレンドを除く必要がある。本節では、測定で得られたデータに補正を施し、真のカウント値を求める操作について述べる。

6.2.1 ダーク成分の引き算

測定データのもつダーク成分を取り除くために、ダークフレームを取得して反射像イメージからの引き算をした。ダークフレームとは、高速シャッターを閉じて X線ビームの照射を止め、反射像イメージの撮像と同じ時間だけ露光して取得したイメージをいう。このイメージはダーク成分のみを含むと考えられるので、HXT-2の性能測定では、各スキャンパスごとにダークフレームを撮像して対応する反射像イメージから差し引いた。

Aperture of the telescopeand positions of beam

6.2 データの補正 67

Fig. 6.2 測定の概念図。ラスタースキャン法を用いて、一つのセグメントごとに測定した。左図のような 3カ所に X線ビームを当てると、中央図のような細長い反射像がそれぞれ得られる。最終的に全てのパスをスキャンすると、右図のような蝶々型の反射像が出来上がる。

ラスタースキャンによる測定の概念図を Fig. 6.2に示した。ラスタースキャンの各スポット（10 mm四方の X線ビームの当たるそれぞれの領域）に X線が当たると、3.4.3項で議論した理由により焦点面では細長い反射像をつくる。最終的には、全てのスポットで得られた反射像を足し算することで、望遠鏡全体に平行な X線を当てて得られる焦点面像を擬似的に再現できる。

6.2 データの補正測定で取得したある生データ（反射像イメージ）は一般的に、式 (6.1)に示すような成分を含んでいる。

I ′(y, z) = c(y, z)× I(y, z) + dark + bias (6.1)

ここで、(y, z)は検出器面上での座標、I ′(y, z)は取得した反射像イメージのカウント値、I(y, z)は真のカウント値、c(y, z)は検出器がピクセルごとにもつ検出効率、darkは暗電流によるノイズ成分、biasは測定値のもつバイアス成分を表す。求めたい真のカウント I(y, z)を取り出すには検出器の検出効率を一様に揃え、ノイズ成分やバイアス成分を差し引かなければならない。性能測定においては dark（ノイズ成分）と bias（バイアス成分）を区別しないので、まとめてダーク成分と呼ぶ。HXT-2の性能測定はラスタースキャンで行うため、最終的に得られる蝶々型の反射像イメージには式 (6.2)に示す

ような成分が含まれる。

I ′add(y, z, t) =522!

i=1

(fi(t)× I ′i(y, z)) (6.2)

ここで、fi(t)は X線ビーム強度の時間変動を表す。ラスタースキャンでの測定には長時間を要するため、ビーム強度の長期的な時間変化の効果が影響する。つまり真のカウント値 I(y, z)を取り出すには、ダーク成分と検出器の感度ムラ補正に加えて、ビーム強度の長期トレンドを除く必要がある。本節では、測定で得られたデータに補正を施し、真のカウント値を求める操作について述べる。

6.2.1 ダーク成分の引き算

測定データのもつダーク成分を取り除くために、ダークフレームを取得して反射像イメージからの引き算をした。ダークフレームとは、高速シャッターを閉じて X線ビームの照射を止め、反射像イメージの撮像と同じ時間だけ露光して取得したイメージをいう。このイメージはダーク成分のみを含むと考えられるので、HXT-2の性能測定では、各スキャンパスごとにダークフレームを撮像して対応する反射像イメージから差し引いた。

Summed image

image at the focal plane(local spot images)

・Pt + C 多層膜・N = 15 + Pt・Γ = 0.5・d = 30 Å

・測定セットアップ上流(DCMと集光ミラー) 下流(サンプル、検出器etc.)

・サンプル・基板：”Super Substrate” ・反射膜：

Pt

Pt

Pt

C

・・・

・Pt + C 多層膜・N = 15 + Pt・Γ = 0.5・d = 30 Å

・測定セットアップ上流(DCMと集光ミラー) 下流(サンプル、検出器etc.)

・サンプル・基板：”Super Substrate” ・反射膜：

Pt

Pt

Pt

C

・・・

9-22

.

2.2 Data Analysis systilt and systwist Fig. 5 shows schematic view to derive systilt and systwist parameters of a subfoil in the mirror file from the spot images. The systilt and systwist parameters were derived from a centroid of each spot image measured at 30 keV because the profile at 30 keV has smallest statistical and systematic uncertainty even for the outer most spot.

×

×

Ideal focus point +image centroid

ΔXs[mm]

ΔYs[mm]

Δθ= ΔXs / f（f: focal length）systilt_pri = Δθ/4, systilt_sec = -Δθ/4Δφ= ΔYs / R (R: radius of subfoil)systwist_pri = systwist_sec = Δφ

R

local spot image

66 第 6章 X線を用いた HXT-2の性能評価

-200

-150

-100

-50

0

50

100

150

200

-200 -150 -100 -50 0 50 100 150 200scan_path

Fig. 6.1 ラスタースキャンのスキャンパス。横軸は Y方向、縦軸は Z方向のステージ移動量を表し、望遠鏡中心を原点としたときの位置を mm 単位で表示した。図中に示した正方形の枠が 10mm 四方のビームを照射する領域を表し、この枠内に示した数字がスキャンパスのシーケンス番号を表す。この図は、Fig. 5.8で定義したホールサイド配置におけるビームのスキャンパスを示す。X線は、紙面の奥から手前に向かって入射する。

プルステージを移動させる。このとき望遠鏡と検出器の位置関係を保つために、サンプルステージと同期させて検出器ステージも移動させる。

この測定方法をラスタースキャン法、あるいは単にラスタースキャンと呼ぶ。Fig. 6.1に、ラスタースキャンにおけるスキャンパスの例を示した。なお、望遠鏡を取り付けているサンプルステージや検出器ステージの稼動域に制限があり、一挙に全セグメントのスキャンができないため、HXT-2の性能測定は一つのセグメントごとに行った。測定するセグメントは常に、Fig. 5.8で定義したホールサイドに配置した。

Fig. 4 Reflectivity measurement at SPring-8 BL-01B Configuration of BL-01B (left) and instrument and sample setup (right)

Fig. 5 Schematic view of the measurement data and determination of misalignment parameters; systilt and systwist.

10-22

.

Scattering Profile for front side of reflectors Aperture is divided into regions defined with the boundaries of multilayer group and sector boundaries, and averaged 1-dimensional profiles of the local spot images corresponding to those regions were derived. Front side reflection profiles of sub-foils of reflectors were obtained by deconvolution using FFT from those 1-dimentional profile with an assumption that primary and secondary reflectors have the same reflection profile. Fig. 7 shows a profile of a spot image and derived single-reflection profile by deconvolution In the current scattering file, the reflection profiles at 30, 40, 50 and 60 keV are implemented. For HXT2, profiles widened by a factor of 1.12 are used to adjust an encircled energy function of the HXT2 to the measured one. For HXT1, adjustment of encircled energy function has not been done.

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R070

R070(R<80)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R090

R090(80<R<100)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R110

R110(100<R<120)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R130

R130(120<R<140)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R150

R150(140<R<160)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R170

R170(160<R<180)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R190

R190(180<R<200)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg1 R210

R210(200<R<220)

ï10 ï5 0 5 10ï10

ï5

0

5

10

Y[m

m]

X[mm]

Seg3 R230

R230(220<R<225)

Fig. 6 distribution of centroids of local spot images in each annulus of aperture for HXT2 Black and red crosses represent positions of centroids of local spot images and weighted mean of them, respectively

11-22

.

Reflectivity Curve The measurement reflectivity of Pt/C multilayer mirror at energies around Pt-L and Pt-K absorption edges is consistent with the reflectivity calculated from the scattering factors of ‘HenkeSskChant’ for interesting range of incident angle and energies of absorption edge (Fig. 8).

• Segment 2, Sector 5, SM901-90430keV

40keV

50keV

60keV

−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]

Scattering angle [arcmin]

SP8Seg2_sm900_sec5_30keV, Data(Black),Dec(R),Conv(G)

−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



• Segment 2, Sector 5, SM3-430keV

40keV

50keV

60keV

−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]


SP8Seg2_sm3−4_sec5_30keV, Data(Black),Dec(R),Conv(G)

−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



• Segment 2, Sector 5, SM630keV

40keV

50keV

60keV

−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



• Segment 2, Sector 5, SM830keV

40keV

50keV

60keV

−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



−20 0 2010−5

10−4

10−3

0.01

0.1

1

10

Prob

abili

ty [/

rad.

]



Fig. 7 1-dimensional profiles for HXT2. Black : local spot image, Red : deconvolved reflection profile, Green: convolved profile derived from the reflection profile.

12-22

.

Fig. 8 Reflectivity of Pt/C multilayer. Top two figures show the reflectivity at energies around Pt-L edges at incident angles of 0.1 and 0.2 degrees; red: model with current optical constants, green: measured data, blue: ratio of model to data. Reflectivity including Pt-K edge at incident angles of 0.04, 0.05, 0.06 and 0.07 degrees are shown in bottom 4 figures; red line: model with current optical constants, black dots: measured data 2.3 Results Summary of the current settings in the CALDB files

Incident angle = 0.2 degIncident angle = 0.1 deg

Red: Model with current optical constantsGreen: DataBlue: Model / Data

Pt-L3

Pt-L2Pt-L1

11 12 13 14 15

1

0.5

Refle

ctiv

ity

IncidentAngle=0.1

11 12 13 14 15

0.9

1

1.1

1.2

Energy [keV]

11 12 13 14 15

1

0.5

Refle

ctiv

ity

IncidentAngle=0.2

11 12 13 14 15

0.9

1

1.1

1.2

Energy [keV]

!"#$% &"#$% #$'#$�(

$)$#

$)#

#

*+,-+./010/2

34+*526+78

$)$9:+5

!"#$% &"#$% #$'#$�(

$)$#

$)#

#

*+,-+./010/2

34+*526+78

$)$!9+5

!"#$% &"#$% #$'#$�(

$)$#

$)#

#

*+,-+./010/2

34+*526+78

$)$'9+5

!"#$% &"#$% #$'#$�(

$)$#

$)#

#

*+,-+./010/2

34+*526+78

$)$%9+5incident angle = 0.04 deg incident angle = 0.05 deg

incident angle = 0.06 deg incident angle = 0.07 deg

Pt-K Pt-K

Pt-K Pt-K

Red line : Model with current optical constants Black : Data

13-22

.

1) Mirror file

i) MIRROR extension • Positions of edge points in cylindrical coordinate

Design values • Misalignment in the direction of tilt and twist

Determined from ground calibration data.

ii) OBSTRUCT extension Design values

iii) SEGMENT extension

Design values

iv) COLLIMATOR extension Design values

v) SURFACE extension

Design values of multilayer structures Roughness derived to fit the energy dependence of the effective area of each multilayer group obtained by the ground calibration.

vi) AZIMUTHALSTRUCT extension : definitions of thermal shield geometry and materials

Design values 2) Scatter file : reflection profiles for front side and back side of reflectors and precollimator

blades Front side of reflectors : derived from spot image profile obtained by ground calibration Back side of reflectors, precollimator blades : profile of back side of Suzaku XRT reflector

3) Reftrans file

Front side of reflectors : calculated by xrtreftable with the current CALDB files Back side of reflectors, precollimator blades : reflectivity tables for back side of reflectors and both side of precollimator blades from Suzaku data blow 20 keV and calculation above 20 keV, and mass absorption coefficient table for the thermal shield and the central cover from calculation with the current absorption coefficients.

4) atomicScattering file

‘HenkeSskChant’ is used for calibration On-axis Effective Area and HPD Fig. 9 and Fig. 10 show on-axis effective area and HPD estimated by xrtraytrace with the current CALDB files in comparison to the values obtained on the ground for HXT1 and HXT2,

14-22

.

respectively. Effective area was derived by accumulating flux within a radius of 4.3’ (16 mm at the focal plane). Discrepancy between the simulated and the measured effective area is smaller than +9/-7 % at energies of 8, 20, 30, 40, 50, 60 and70 keV.

On-axis EEF and PSF Fig. 11 and Fig. 12 show EEF and PSF of HXT1 and HXT2, respectively. EEF and PSF were calculated from a image normalized so that a flux within a radius of 6.24’ is equal 1 for both of

1

1.5

2

2.5

HPD

[arc

min

]

comp2_hpd_tlps_twps_sc1_full.qdp 1117b3,1203Experiment

Simulation

0 20 40 60 80

0.9

1

1.1

Sim

/Exp

Energy [keV]

10

100

1000

Effe

ctiv

e A

rea

[cm

2 ]

comp2_ea_tlps_twps_sc1_full.qdp 1117b3,1203

SimulationExperiment

0 20 40 60 80

0.9

1

1.1

Ratio

Energy [keV]

Fig. 9 Effective area (left) and HPD (right) of the HXT-1 (full telecope without the thermal shields)

10

100

1000

Effe

ctiv

e A

rea

[cm

2 ]

comp2_ea_tlps_twps_sc1_full.qdp,1117a,1203mod3

SimulationExperiment

0 20 40 60 80

0.9

1

1.1

Ratio

Energy [keV]

1

1.5

2

2.5

HPD

[arc

min

]

comp2_hpd_tlps_twps_sc1_full.qdp(150714a,3)Experiment

Simulation

0 20 40 60 80

0.9

1

1.1

Sim

/Exp

Energy [keV]

Fig. 10 Effective Area (left) and HPD (right) of the HXT-2 (full telescope without the thermal shields)

15-22

.

simulated and measured data, and PSF is azimuthally averaged. The simulated EEF is in good agreement with the measured one EEFs as the difference is smaller than 5 % at a radius larger than 2 arcmin. For PSF, discrepancy is less than 20 % within a radius of 4 arcmin. At a radius > 4 arcmin, simulated PSF is 20 - 40 % fainter than the measured one.

0

0.5

1

EEF

comp2_eef_tlps_twps_sc1_e50_full.qdp,1117b3,1203

ExperimentSimulation

0 2 4 6

0.91

1.11.2

sim/d

ata

Radius [arcmin]

0

0.5

1

EEF



0 2 4 6

0.91

1.11.2

sim/d

ata

Radius [arcmin]

10−5

10−4

10−3

0.01

0.1

1

10

Brig

htne

ss [a

rb.]

comp2_psfbin10_tlps_twps_sc1_e50_full.qdp,1117b3,1203


0 2 4 60

0.51

1.52

sim/d

ata

Radius [arcmin]

10−5

10−4

10−3

0.01

0.1

1

10

Brig

htne

ss [a

rb.]



0 2 4 60

0.51

1.52

sim/d

ata

Radius [arcmin]

0

0.5

1

EEF



0 2 4 6

0.91

1.11.2

sim/d

ata

Radius [arcmin]

10−5

10−4

10−3

0.01

0.1

1

10

Brig

htne

ss [a

rb.]



0 2 4 60

0.51

1.52

sim/d

ata

Radius [arcmin]

Fig. 11 EEF (left) and PSF (right) of HXT-1 at 30 keV (top), 50 keV (middle), 70 keV (bottom)

16-22

.

The Crab pulsar EEF The Encircled Energy Function (EEF) was also measured from inflight Hitomi data. For the Crab pulsar image in 5-80 keV band the EEF was generated by excluding a contribution from an extended emission region from the Crab nebula by subtracting an image extracted in the off-pulse phase and normalized with the dead-time corrected exposure time. Fig. 12(b) shows the systematic difference between the observed and simulated EEF. As shown in Fig. 12(b), while the simulated EEF agrees well with the EEF measured on the ground, the

0

0.5

1EE

F

comp2_eef_tlps_twps_sc1_e30_full.qdp(1117a,1203)


0 2 4 6

0.91

1.11.2

sim/d

ata

Radius [arcmin]

0

0.5

1

EEF



0 2 4 6

0.91

1.11.2

sim/d

ata

Radius [arcmin]

0

0.5

1

EEF



0 2 4 6

0.91

1.11.2

sim/d

ata

Radius [arcmin]

10−5

10−4

10−3

0.01

0.1

1

10

Brig

htne

ss [a

rb.]

comp2_psfbin10_tlps_twps_sc1_e30_full.qdp(1117a,1203)


0 2 4 60

0.51

1.52

sim/d

ata

Radius [arcmin]

10−5

10−4

10−3

0.01

0.1

1

10

Brig

htne

ss [a

rb.]



0 2 4 60

0.51

1.52

sim/d

ata

Radius [arcmin]

10−5

10−4

10−3

0.01

0.1

1

10

Brig

htne

ss [a

rb.]



0 2 4 60

0.51

1.52

sim/d

ata

Radius [arcmin]

Fig. 12 EEF (left) and PSF (right) of HXT-2 at 30 keV (top), 50 keV (middle), 70 keV (bottom)

17-22

.

Crab EEF is systematically larger than the simulated EEF. The difference depends on a radius and it is larger than 5 % at a radius smaller than 2 arcmin. The ratio plotted in Fig. 12(b) represents the systematic error of the effective area as a function of extraction radius. When extracting the spectrum within a radius smaller than 2 arcmin, the raytracing code xrtraytrace and thus the arf generator give ~5 % smaller effective area than that expected from the Crab EEF using the current mirror and detector calibrations. Therefore large extraction radius is recommended to avoid large systematic error of effective area and consequently the source flux estimation.

Fig. 12b Encircled Energy Function normalized at 4 arcmin (Upper panel, black points : the Crab pulsar in 8-80 keV, red solid line : simulated EEF at 30 keV, green solid line: simulated EEF at 50 keV) and ratio of observed to simulated EEF (bottom, red points: the Crab pulsar in 5-80 keV / simulated at 30 keV, green points: the Crab pulsar in 5-80 keV / simulated at 50 keV, red dashed line: on-ground EEF at 30 keV / simulated EEF at 30 keV, green dashed line: on-ground EEF at 50 keV / simulated EEF at 50 keV) . Vignetting Function of each segment at 50 keV Because vignetting function was measured at 50 keV in two orthogonal directions (θy and θz; see ) for each segment independently, the vignetting function for the full telescope configuration has not been measured in any direction. In Fig. 13 and Fig. 14, the simulated vignetting functions of HXT1 and HXT2 were compared with measured ones by segment in two measured directions, respectively. Measured and simulated vignetting functions show in these figures are normalized by the peak effective area and its peak is aligned to offaxis angle = 0 arcmin. When offaxis angle < 2’, simulated vignetting function is in quite good agreement with measured one as the difference is less than 5 %. Even for offaxis angle > 2’, discrepancy is expected to be smaller than 10 %.

0

0.5

1

EEF(

norm

aliz

ed a

t 4’)

hxt1_e30_50_merge.qdp

0 1 2 3 4

0.9

1

1.1

Ratio

Radius [arcmin]

0

0.5

1

EEF(

norm

aliz

ed a

t 4’)

hxt2_e30_50_merge.qdp

0 1 2 3 4

0.9

1

1.1Ra

tio

Radius [arcmin]

Points: The Crab EEF / Simulated EEF Dashed line: on-ground EEF / simulated EEF

Points: The Crab EEF / Simulated EEF Dashed line: on-ground EEF / simulated EEF

Points: The Crab EEF Solid line: simulated EEF (red: 30keV, green: 50 keV)

Points: The Crab EEF Solid line: simulated EEF (red: 30keV, green: 50 keV)

18-22

.

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea

comp2_vig_tlps_twps_sc1_e50_seg2_tz_norm_shift_nolink.qdpSimulationExperiment 151117b3,1203Peak aligned

−5 0 5

0.8

1

1.2

Ratio

offaxis angle [arcmin]

θz

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θz

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea

comp2_vig_tlps_twps_sc1_e50_seg2_ty_norm_shift_nolink.qdpSimulationExperiment 151117b3,1203Peak aligned

−5 0 5

0.8

1

1.2

Ratio


θy

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θy

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θz

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θy

Fig. 13 Vignetting function of the segment 1 (top), the segment 2 (middle), segment 3 (bottom) of HXT-1 in θy (left) and θz (right) directions at 50 keV (segment number defined in the telescope definition file). Segment and rotation axis are shown in the inset figure.

19-22

.

Fig. 14 Vignetting function of the segment 1 (top), the segment 2 (middle), segment 3 (bottom) of HXT-2 in θy (left) and θz (right) directions at 50 keV (segment number defined in the telescope definition file). Segment and rotation axis are shown in the inset figure.

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea

comp2_vig_tlps_twps_sc1_e50_seg3_ty_norm_shift.qdpSimulationExperiment 1117a,1203Peak aligned

−5 0 5

0.8

1

1.2

Ratio


θy

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea

comp2_vig_tlps_twps_sc1_e50_seg3_tz_norm_shift.qdpSimulationExperiment 1117a,1203Peak aligned

−5 0 5

0.8

1

1.2

Ratio


θz

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θy

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θy

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θz

0.1

1

0.2

0.5

Nor

mal

ized

Effe

ctiv

e A

rea


−5 0 5

0.8

1

1.2

Ratio


θz

20-22

.

Analytic function of Vignetting For convenience to calculate ARF, analytic function representing the azimuthally averaged vignetting function of HXTs were obtained. Azimuthally averaged vignetting function was derived from simulated 2 dimensional vignetting maps at energies of 10 to 70 keV with 10 keV pitch with xrtraytrace and the current CALDB files, and fitted with the formula of lorentizan + constant;

Here. E and θ are a energy in keV and an off axis angle in arcmin. The coefficients LN(E) and LW(E) are the functions of energy as followings, For HXT1, For HXT2, Discrepancy between analytic function and simulated one are +/-5 % at E < 60 keV and +/-10 % at E ≥ 60 keV for HXT1, and +/- 5 % for HXT2 as shown in Fig. 15. We should note that the difference shown here is not between measured data and the function but also between the xrtraytrace result and the function, and the analytic function is only applicable within a off axis angle of 7 arcmin.

Fig. 15 Azimuthal averaged vignetting curve from the analytic function and raytracing simulation with xrtraytrace and the current CALDB files for HXT1 (left) and HXT2 (right). Bottom figures show the ratio of analytic function to xrtraytrace result. Dots and lines represent xrtraytrace and analytic function, respectively. Colors express different energies of 10 (black), 20 (red), 30 (green), 40 (blue), 50 (light blue), 60 (magenta) and 70 (yellow) keV.

0.1

1

0.2

0.5

Vig

netti

ng

10−70keV, lor+co, HXT1

0 2 4 6

0.9

1

1.1

Mod

el/D

ata

Off−axis angle [arcmin]

0.1

1

0.2

0.5

Vig

netti

ng

10−70keV, lor+co, HXT2

0 2 4 6

0.9

1

1.1

Mod

el/D

ata

Off−axis angle [arcmin]

LW (E) = 10.224 � 1.148 exp(E/39.155)

LN(E) = �1.955 � 10�4E + 0.95755

LW (E) = 14.865 � 5.323 exp(E/96.094)

LN(E) = �3.064 � 10�4E + 0.96278

21-22

.

2.4 Final remarks The accuracy of simulation with the current CALDB files is estimated from the comparison with the ground calibration data taken at SPring-8; discrepancy of on-axis effective area between the simulation and the measurement is +/-9 % at energies of 8, 20, 30, 40, 50, 60 and70 keV, that of the EEF is 5 % or less at a radius larger than 2 arcmin, that of PSF is less than 20 % within a radius of 4 arcmin and at a radius > 4 arcmin, simulated PSF is 20 - 40 % fainter than the measured one. 3 Release CALDB 20161122

Filename Valid data Release date CALDB Vrs

Comments

ah_hx1_mirror_20140101v001.fits 2014-01-01 20161122 005 Mirror ah_hx2_mirror_20140101v001.fits 2014-01-01 20161122 005 Mirror ah_hx1_reftrans_20140101v001.fits 2014-01-01 20161122 005 Reftrans ah_hx2_reftrans_20140101v001.fits 2014-01-01 20161122 005 Reftrans ah_hx1_auxtran_20140101v001.fits 2014-01-01 20161122 005 Fudge factor ah_hx2_auxtran_20140101v001.fits 2014-01-01 20161122 005 Fudge factor

New mirror definition files are released for HXT-1 and HXT-2 but the only change was the addition of a new keyword (TELFPROT) that enables the updated xrtraytrace code to apply a rotation to the PSF with respect to the focal plane in order to match the output of the raytracing to the physical coordinate systems. Also released are new reflectivity files for HXT-1 and HXT-2. The only change in these files is the addition of data at several new energies to improve the energy grid. The final two files in the above table are new and (optionally) implement “fudge factors” to modify the effective area in the ARF and spectral response (RSP) files generated by aharfgen and hxirspeffimg. The data in these files is described below. 3.1 Data description The “auxiliary transmission” factors (or fudge factors) that are used to modify the effective area generated by aharfgen and hxirspeffimg for HXI1 and HXI2 are adjustments to the effective area that are intended to account for the difference between the ground-based telescope effective area measurements and the telescope effective areas generated by the raytracing code xrtraytrace. The effective area produced by the raytracing code is controlled by the data in the mirror definition files, the reflectivity files, and the scattering files. Although the data in these files were fine-tuned to reproduce the ground-based effective areas as closely as possible by adjusting physical parameters relating to the physical properties of the telescopes, residuals of up to ~7% between the ground-based measurements and the raytracing predictions remained. These residual discrepancies have not been resolved in terms of identifying their physical origin. The CALDB files described here can be used optionally with aharfgen and hxirspeffimg to include the correction to the effective area in the ARF. Since the discrepancies are not negligible in the cases

22-22

.

of observing bright sources such as the Crab, the fudge factors are introduced to reproduce the effective area obtained from the on-ground calibration.

Fudge factor is a spline function which smoothly connects the ratios of measured to simulated effective area at 8, 30, 40, 50, 60 and 70 keV (and 20 keV only for HXT2) and additional knots to make the function converge with the raytracing output at the lowest and highest emergy as shown in Fig. 16. 3.2 Data analysis Tools locally developed by the HXT team have been used to analyze the ground calibration data and estimate effective areas. To generate fudge factor table by spline interpolation, a perl script “mk_arf_fudge160928.pl” locally developed has been used. 3.3 Final remarks The fudge factors introduced to reduce a discrepancy between the effective area measured on the ground and that of raytracing output.

0 50 100

0.9

1

1.1

Dat

a/Si

m

Energy [keV]

HXT1

Data/Sim(no Fudge)Fudge160928

additional knots

0 50 100

0.9

1

1.1

Dat

a/Si

mEnergy [keV]

HXT2

Data/Sim(no Fudge)Fudge160928

knots for shaping

Fig. 16 Fudge factor as a function of energies (red solid line). Open circles with an error bar represent a ratio of effective area measured on the ground to that of the raytracing output with no fudge factor. Filled circles are additional knots to make the function smoothly converge with the raytracing output at the lowest and highest energy.

Documents

JAXA / GSFC - NASA · 2016-12-23 · ASTRO-H INSTRUMENT CALIBRATION REPORT Mirror Definition File ASTH-HXT-CALDB-MIRROR Version 0.2 December 22, 2016 JAXA / GSFC Prepared by: A. Furuzawa,