BEAMLINES SOPHISTICATED SYSTEMS CONSTRUCTED FROM SIMPLE BUILDING - PowerPoint PPT Presentation

BEAMLINES SOPHISTICATED SYSTEMS CONSTRUCTED FROM SIMPLE BUILDING BLOCKS An introduction to beamline engineering Author - Title (Footer) 2

BEAMLINES SOPHISTICATED SYSTEMS CONSTRUCTED FROM SIMPLE BUILDING BLOCKS An introduction to beamline engineering Outline What are we dealing with? Basic beamline functions and generic solutions. Heatload Management Vacuum Supports Positioning Author - Title (Footer) 3

What comes out of the Front End (looking towards the machine) Radiation from Dipole magnet Undulator beam Radiation from Dipole magnet downstream of straight section upstream of straight section Author - Title (Footer) 4

What is delivered to the beamline (high power front end) Aperture in Front End defining the beam. 4mm diameter diamond window Author - Title (Footer) 5

100 120 20 40 60 80 0.002 0 0.0015 0.001 100-110 0.0005 90-100 80-90 70-80 0 60-70 50-60 -0.0005 40-50 30-40 -0.001 20-30 10-20 -0.0015 0-10 -0.002 0.002 0.0015 0.001 0.0005 0 -0.0005 -0.001 -0.0015 -0.002 120 Power frofile for a 100 single u27 undulator 80 60 @200mA at the primary 40 slits 20 0 1 2 3 4 5 6 7 8 9 Author - Title (Footer) 6

Laser welder 8 KW can welder metal tubes seams at 60m/min Power Comparisons New upgrade beamlines UPBL11 8 metres of undulator u27 gap 1mm gives max power density 385 Watts/mm2 Total power through the Front End approx 4 kWatts Author - Title (Footer) 7

Outline What are we dealing with? Basic beamline functions and generic solutions. Heatload Management Vacuum How do we process these beams? Supports Positioning Slits/Absorbers Windows Optics Author - Title (Footer) 8

How do we process these beams? Slits 4 Independent blades each capable of intersecting the entire beam. Power in Q Fluid temperature Proportional to flow rate and Q Choose Wall temperature Maximum temperature Copper (high thermal Dependant on exchange conductivity) coefficient (Flowrate) Proportional to k thermal conductivity Water cooling (turbulent flow) Proportional to Q Dependant on geometry Author - Title (Footer) 9

How can this fail? Simple approach: The surface should not melt! More conservative approach: The thermal induced stress should not exceed 2 x thermal yield strength of the material. Choice of ―Glidcop‖ a dispersion strengthened copper which maintains high yield stress even after brazing Melting temp o C Choice of material Conductivity W/mK With optimized cooling geometry Diamond 3550 900-2300 gives a maximum power Copper 1084 401 absorption of 50W/mm length of Aluminium 660 237 absorber Stainless Steel 1400 12-45 Author - Title (Footer) 10

Initial Primary Slit design q Q W/mm Power absrbed per mm of absorber = Qsin q If we have power densities of Why was it so big? 1000 W/mm then q should z Present High Power Slits be less than 2.8 degrees y z y Grazing angle 1.8 degrees Author - Title (Footer) 11

How do we process these beams? WINDOWS Window, attenuators and CRLs can be treated the same. h t In the central zone this can be approximated to a 1 dimensional problem Assuming copper at constant temperature (20C) Q absorbed = -kAdT/dx  Q abs/mm = k x t x D T x 2/h  D T = h Q abs/mm /2kt Example a.0.3mm diamond in u27@11mm absorbs 28W/mm if h =6mm then temperature = 140 o C Materials and thicknesses Example b 0.3mm aluminum in u27@11mm absorbs 41W/mm if h =6mm then temperature = 1180oC Author - Title (Footer) 12

In real conditions • Beam is not linear approx 3mm horizontally x 1mm vertically. • Copper cooling block is not constant temperature • Material is not isotropic HP attenuators Author - Title (Footer) 13

Typical attenuator foils that are now used CVD diamond Pyrocarbon Pure aluminium CVD diamond coated with high z material Not to forget our initial problem! Author - Title (Footer) 14

How do we process these beams? OPTICS • Monochromators • Mirrors The difference: size of the projected beam on the optical surface. Monochromator Mirror Angle of incidence (Bragg Angle) 4-70 degrees Angle of incidence (Bragg Angle) 0.1 to 0.3 degrees Typical Footprint 3mm x 3mm Typical Footprint 3mm x 500mm Power density up to 100 W/mm2 Power density up to 1 W/mm2 Total power up to 300W Total power up to 800W Author - Title (Footer) 15

How do we process these beams? OPTICS Mirrors Mirror cooled by In/Ga baths. In/Ga is a ―safe‖ mercury. Liquid metal for good flexible thermal contact Author - Title (Footer) 16

How do we process these beams? OPTICS Results temperature along mirror 45.00 40.00 temperature deg C 35.00 200mA Secondary slit 160mA 30.00 160mA width corr 25.00 Primary Slits 20.00 0 100 200 300 400 500 distance along mirror (mm) If the mirror is inserted in the white Deformation of mirror End effect outside useful beam beam part will be reflected and 4.50E-06 part will be absorbed. The absorbed beam will heat up the 4.00E-06 surface of the mirror and also the 3.50E-06 bulk of the mirror. 200mA deformation (m) 3.00E-06 corrected 200mA The temperature of the surface 2.50E-06 160mA can be calculated and varies 160mA corrrected 200mA 2.00E-06 according to above graph. It will 160mA width corr 1.50E-06 vary with different currents in the 160mA width corr 200mA 1.00E-06 machine.. 5.00E-07 From these temperatures the 0.00E+00 deformation of the surface can be 0 50 100 150 200 250 300 350 400 450 calculated and corrected using a distance along mirror bender. Author - Title (Footer) 17

The New ESRF Generic Mirror Solutions Final choice for ESRF ID24_MH1 — Specific horizontally deflecting mirrors ID24 ―Smart‖ section mirrors Water cooling on top of side Clamps for cooling absorbers and Indium foil interface T Mairs et al Upgrading Beamline Performance : Ultra Stable Mirror Developments at ESRF 18

How do we process these beams? OPTICS Monochromators Optimised cooling as seen before. Author - Title (Footer) 19

Monochromators Why liquid nitrogen Thermal conductivity (w/mK) vs temperature 1800 1600 1400 1200 1000 800 600 400 200 0 Aim is to keep this region as flat as possible 70 120 170 220 270 320 temperature (K) Author - Title (Footer) 20

After the monochromator the heatload problem disappears PHEW! But we still have Outline What are we dealing with? Basic beamline functions and generic solutions. Heatload Management Vacuum Supports Positioning Author - Title (Footer) 21

VACUUM Vacuum is not the specialisation of the mechanical engineers working on beamlines, but vacuum chambers are used everywhere. Some figures: ESRF subsystem length vacuum chambers vacuum level 10 -8 mbar Booster 300 metres 10 -11 mbar Storage Ring 32 cells x 26m =844 metres 10 -9 mbar Front Ends 47 x 20m = 940 metres 47 x 20-100m = 1500-2500metres 10 -6 – 10 -10 mbar Beamlines Vacuum Regimes Generally beamlines operate in Rough Vacuum: Atm (1000mbar) - 10-2mbar the high vacuum region. Process Vacuum: 10-2mbar - 10-4mbar But why? High Vacuum: 10-5mbar - 10-9mbar Ultra-High Vacuum: < 10-9mbar Author - Title (Footer) 22

Beamline Vacuum requirements Maximum transmission of photons Vacuum level transmission for 1 metre at energy mbar 2keV 4.5keV 7keV 12keV 20keV 1000 0% 0.5% 25% 76% 93% 100 0.2% 59% 87% 97% 99% 10 55% 95% 98% 99% 100% 1 94% 99% 99% 100% 100% 0.1 99% 100% 100% 100% 100% Cleanliness Surfaces are free from contaminents -to avoid damage i.e. mirrors, crystals etc -for the science of the surface. Practical considerations. -need to minimise vibrations (no mechanical movements  ion pumps) -minimise maintenance (ion pump lifetime  vacuum <10 -7 mbar) Author - Title (Footer) 23

Standardisation at ESRF Decision taken in 1990 to use CF flanges with copper gaskets Conflat flanges Gaskets Bolts Stores 316LN Stores machined Stores silver plated copper Beamlines 304L is OK For rough pumping Author - Title (Footer) 24

Some chambers at ESRF Author - Title (Footer) 25

Outline What are we dealing with? Basic beamline functions and generic solutions. Heatload Management Vacuum Supports Positioning One of the most common jobs to do for our colleagues in MEG ― I have got to go and draw a support!‖ Not an interesting job but very important for the performance of the beamline. Author - Title (Footer) 26

Why are the supports so important? Sample Optics Machine Stability of the beam from the machine is Microbeam: ideally given as 10% of its neither the beam or the We do not want to amplify the divergence sample should move instabilities in the beam. It is more than 10% of its size important that all the optical I.e.<1-2 m m elements and the sample are correctly supported Author - Title (Footer) 27

What do we mean by a good support? • ESRF site already shows movements of the order of 1 micron. Weekends and nights are better. • The support should not amplify these levels. • The support should not resonate with ―normal‖ driving sources. Water flow, LN2 flow, electronic fans, chillers etc. •―Support‖ is not just the vacuum chamber but is also the internal mechanics.

The ideal support? Not an optimised support Distance floor Distance to instrument instrument inposed to support should be minimised