**6.1 In-situ X-ray mirror thermal distortion measurement**

Mirror distortion under high heat load has been recognized as a serious problem for thirdgeneration synchrotron light sources, as well as for the first- and second-generation sources operating under conditions of reduced emittance, high current and with novel insertion devices. Efforts have been made at reducing mirror distortion by the use of high thermal conductivity or low expansion materials, cryogenic cooling, enhanced heat exchangers, jet cooling, and other means. Model calculations based on finite element design codes are used to predict distortion theoretically. However, the performance of any mirror will ultimately be determined experimentally when fully illuminated by the high power synchrotron beam, because the practical boundary conditions usually differ from the ideal theoretical conditions. An in-situ distortion test is then very useful for confirming the theoretical calculations. A precise measuring method to detect the in-situ distortion profile of a high heat load mirror for synchrotron radiation by use of the penta-prism Long Trace Profiler was developed in 1995 in Italy. A schematic diagram of the heat-load measurement equipment of the *in situ* LTP is shown in Fig. 21 (Qian et al., 1997, 1995). The optical head of the LTP II profiler is mounted horizontally on an optical table (TB). The first penta-prism (PT1) scans the sampling beam (SB) along the mirror under test (MUT) by use of a mechanical translation slide (MS) fixed to TB with a 250-mm travel length. The reference beam (RB) is directed onto a fixed spot on the MUT by a mirror (M1) and another pentaprism (PT2). This spot is located along the center of the length of the mirror but displaced by

Angular calibrations of profilers can be made at national standards bureaus. However, for researchers involved in precision angular R&D projects, it is necessary to check the nanoradian accuracy of a profiler frequently, and for calibration at remote sites, a low cost

In the case of sine bar and tangent bar systems for large angle measurements, the contact surface in using mechanical length gauge or reflection mirror surface in interferometric

Most measurements for precision optics are made in controlled environment in order to verify compliance with specifications. However, the actual in-situ use of the optical components could be very different from the laboratory condition. Beam quality and position can be affected by temperature instability, distortion under high vacuum condition, a noisy vibration environment, and thermal distortion due to absorption of high power beams from synchrotron and FEL sources. In these cases, on-line figure measurement of bending mirrors and adaptive optics is highly desirable. The manufacturing of large astronomical optical components with nanometer accuracy requires in-situ testing without removing the optics from the polishing machine. These insitu situations present challenges to the metrology. Most of these metrologies are having

Owing to the restricted conditions for measurements on in-situ optics, very few

Mirror distortion under high heat load has been recognized as a serious problem for thirdgeneration synchrotron light sources, as well as for the first- and second-generation sources operating under conditions of reduced emittance, high current and with novel insertion devices. Efforts have been made at reducing mirror distortion by the use of high thermal conductivity or low expansion materials, cryogenic cooling, enhanced heat exchangers, jet cooling, and other means. Model calculations based on finite element design codes are used to predict distortion theoretically. However, the performance of any mirror will ultimately be determined experimentally when fully illuminated by the high power synchrotron beam, because the practical boundary conditions usually differ from the ideal theoretical conditions. An in-situ distortion test is then very useful for confirming the theoretical calculations. A precise measuring method to detect the in-situ distortion profile of a high heat load mirror for synchrotron radiation by use of the penta-prism Long Trace Profiler was developed in 1995 in Italy. A schematic diagram of the heat-load measurement equipment of the *in situ* LTP is shown in Fig. 21 (Qian et al., 1997, 1995). The optical head of the LTP II profiler is mounted horizontally on an optical table (TB). The first penta-prism (PT1) scans the sampling beam (SB) along the mirror under test (MUT) by use of a mechanical translation slide (MS) fixed to TB with a 250-mm travel length. The reference beam (RB) is directed onto a fixed spot on the MUT by a mirror (M1) and another pentaprism (PT2). This spot is located along the center of the length of the mirror but displaced by

measurements have been made, even though they are very important.

**6.1 In-situ X-ray mirror thermal distortion measurement** 

distance sensor is tilted. This will degrade the measurement accuracy considerably.

precision angle calibration device is desired.

increasing demands on nanometer level.

**6. Precision metrology of in-situ and at-wavelength** 

15 mm transversely toward the edge. For symmetry reasons this point should not have a tangential slope variation component even if the mirror is subject to a high heat load.

Fig. 21. Schematic of the in situ LTP test on beam line

A maximum distortion of 0.47 µm over a length of 180 mm was measured for an internally water-cooled mirror on an undulator beam line at ELETTRA while exposed to a total emitted power of 600 watts (Fig. 22). For this measurement, the configuration with all of the equipment external to the vacuum chamber was used. The experiment has an accuracy and repeatability of 40 nm.

Fig. 22. In situ height distortion profiles of a synchrotron radiation mirror under a high heat load, as measured with the ppLTP: (a) Total power, 600W; undulator gap, 30mm; current, 181mA; energy, 2Gev. (b) Total power, 360W; undulator gap, 40mm; current, 187mA; energy, 2Gev. (c) Total power, 150W; undulator gap, 60mm; current, 224mA; energy, 2Gev. (d) No synchrotron beam on, thus corresponding to the test repeatability (<0.04m, peak to valley)

The second thermal distortion test was done at The Advanced Photon Source (APS) at Argonne National Laboratory on the second mirror of SRI-CAT 2-ID-C beam line in 1997 (Takacs et al., 1998). The in-situ LTP scanned the central 90 mm of the 200 mm long mirror through a vacuum window while the mirror was subjected to heat loading from the synchrotron beam. We measured about 200nm distortion in 90 mm when the X-ray beam was switched on. This distortion was not predicted by the finite element (FE) thermal calculations. By use of adjustable and movable slits upstream of the beam line, the illuminating beam spot position on mirror could be shifted. The aperture width was set to 0.5 mm, which corresponds to a 25 mm beam width on the surface. The very attractive feature of this test is that we can find the thermal bump produced by SR beam with great sensitivity (Fig. 23). The 200nm bump with a 20 rad slope followed the slits center shift movement. The results of this preliminary test indicate that after improvements in instrument stability, it should be able to measure 10nm thermal distortion with ease. This can be very powerful tool for nanometer spot mirror distortion test.

Fig. 23. Thermal bump produced on surface with 98mA beam current. Adjustable aperture is set at 0.5 mm H x 4.0 mm V, corresponding to a 25 mm wide beam on the surface. The aperture center was offset from the normal 0.0 mm position and the measurements show the corresponding lateral shift of the thermal bump. Unit in legend are mm.

#### **6.2 In-situ precision metrology at optical workshop**

In 2001, the portable LTP (PTLTP) was proposed for in-situ measurement at optical workshop to test optics during the final polishing step (Fig. 24) (Qian & Takacs, 2001). The portable LTP has a stationary optical head fixed to polishing machine. The optics to be polished, for example a cylinder mirror, is scanned by a moving table in order to scan the test beam over the mirror. A small mirror is fixed on the moving table in order to test the moving pitch error through a penta-prism.

An actual on-machine measurement has been made on measuring the slope and form errors of long cylindrical mirrors with optical tolerance precision and accuracy at RIKEN in Japan (Moriyasu et al., 2004). Cylindrical surfaces can be measured completely by steering the beam up the sides of the cylinder with a rotational mirror. To measure complete cylindrical surfaces, the direction of the beam to the measured object can be steered by controlling the angle of the flat mirror on the rotation table. In this way, a 2-D surface profile can be obtained.

synchrotron beam. We measured about 200nm distortion in 90 mm when the X-ray beam was switched on. This distortion was not predicted by the finite element (FE) thermal calculations. By use of adjustable and movable slits upstream of the beam line, the illuminating beam spot position on mirror could be shifted. The aperture width was set to 0.5 mm, which corresponds to a 25 mm beam width on the surface. The very attractive feature of this test is that we can find the thermal bump produced by SR beam with great sensitivity (Fig. 23). The 200nm bump with a 20 rad slope followed the slits center shift movement. The results of this preliminary test indicate that after improvements in instrument stability, it should be able to measure 10nm thermal distortion with ease. This

Fig. 23. Thermal bump produced on surface with 98mA beam current. Adjustable aperture is set at 0.5 mm H x 4.0 mm V, corresponding to a 25 mm wide beam on the surface. The aperture center was offset from the normal 0.0 mm position and the measurements show the

In 2001, the portable LTP (PTLTP) was proposed for in-situ measurement at optical workshop to test optics during the final polishing step (Fig. 24) (Qian & Takacs, 2001). The portable LTP has a stationary optical head fixed to polishing machine. The optics to be polished, for example a cylinder mirror, is scanned by a moving table in order to scan the test beam over the mirror. A small mirror is fixed on the moving table in order to test the

An actual on-machine measurement has been made on measuring the slope and form errors of long cylindrical mirrors with optical tolerance precision and accuracy at RIKEN in Japan (Moriyasu et al., 2004). Cylindrical surfaces can be measured completely by steering the beam up the sides of the cylinder with a rotational mirror. To measure complete cylindrical surfaces, the direction of the beam to the measured object can be steered by controlling the angle of the flat mirror on the rotation table. In this way, a 2-D surface profile can be

corresponding lateral shift of the thermal bump. Unit in legend are mm.

**6.2 In-situ precision metrology at optical workshop** 

moving pitch error through a penta-prism.

obtained.

can be very powerful tool for nanometer spot mirror distortion test.

Fig. 24. Proposed in situ test at workshop in 2001
