APRIL 01:

Mirror washing. Iris reflectometry and BRDF data before/after wash was respectively:

• 80.1/91.1% (470nm), 79.5/90.9% (530nm), 76.9/87.2% (650nm), 74.8/86.2% (800nm)
• 7.4/0.5% (+20deg), 6.2/0.6% (-15deg), 1.1/0.05% (-45deg)

Thus a spectacular 11% net gain in reflectivity. Only about 1.5% was not recovered compared to the last washing results. It also shows a 2.2% reflectivity loss per month (averaged over the last 6 months), which is about 3.5 times higher than usual!

NOVEMBER 00:

Mirror washing. Iris reflectometry and BRDF data before/after wash was respectively:

• 87.2/92.6% (470nm), 87.5/93.1% (530nm), 83.4/89.9% (650nm), 83.1/89.0% (800nm)
• 2.4/0.8% (+20deg), 2.5/1.1% (-15deg), 0.4/0.09% (-45deg)

It was very successful, bringing the mirror back close to a fresh aluminium performance.

We used isopropyl alcohol at the end of the rinsing but this did leave a few ugly whitish marks that we could not recover with another soap wash. Rancourt explains (Optical Thin Films, User's handbook, McGraw Hill 1987) that the 'evaporation of the solvent on the surface beeing cleaned will reduce the temperature locally and may cause the moisture in the air to condense in the vicinity of the droplet, leaving a mark when water dries' (p175). This might be what happened. Since then, we do not use alcohol anymore in our washings... We also identified thousands of pinholes in the coating by shining light from underneath the mirror, sign that the mirror was dusty when last put in the coating chamber.

APRIL 00:

We washed M1. The following graph show reflectivity and scatter as we found the mirror after 10 months (without cleaning it at all), after a CO2 cleaning, and finally after washing. Last aluminizing data is also plotted for reference. Final state of the mirror is actually better than after aluminizing, a surprise also noticed by Giordano at the NTT using Collodion peeling films (Wilson R, Reflecting telescope optics II, A&A library 1999, p 444). All measurements made with the Minolta.

M1 reflectivity and scatter [2]

Between June 97 and June 99, M1 reflectivity as decreased by 7.8% per year at 400nm and 7.2 per year at 700nm, while scatter has increased by 3.8% per year at 400nm and 5.3% per year at 700nm. Between June 99 and April 00, M1 reflectivity and scatter have varied according to the same trend as the previous period, which is a monthly variation of: R-0.68% at 400nm and R-0.73% at 700nm, SC+0.36% at 400nm and SC+0.52% at 700nm. In theory, scattered light dependence on wavelength is given by TIS=(4.pi.d/l)^2 where TIS is the total integrated scatter measured in 2pi str, d is the roughness height and l is the wavelength. Therefore, scattered light should be lower in the nearIR than in the blue. Particle size might modify that law if they are about the size of the wavelength.

JUNE 99:

We aluminized both mirrors. Last time was: 6June97 for M1 and 9Dic97 for M2. See the following plots. In the text below, we give the average number (2 points sampled each time) at 550nm for a rapid evaluation of, respectively, the reflectivity and the scattered light measured with the Minolta.

• • M1: was very dirty visually, covered with dust (77.0% and 10.80%). We noted the presence of lots of particles longer than 0.5mm, even a few (10) >1mm. First we cleaned it with CO2 (86.7% and 4.40%). Then we washed it with Orvus soap (contact with natural sponge), rinsed with bi-distilled water and dried with dry nitrogen (91.0% and 1.10%). The result was spectacular, close to a fresh aluminium

         Finally, we aluminized (91.9% and 1.10%).

    M1 reflectivity [3] and M1 scattering [4].

    From the reflectivity of fresh aluminizing (91.9%) and after CO2 cleaning (86.7%), we can estimate the monthly loss if the mirror had been regularly CO2 cleaned: (91.9-86.7) / 24 = 0.22% which is in agreement with the data measured at the 1.5m with regular cleaning.      We conclude that CO2 cleaning slows the natural dust deposition process by a factor of about 3.

  • M2: was quite clean after removal from the telescope (88.1% and 1.80%). Cleaning it with C02 didn't make any difference (87.9% and 1.90%). Aluminizing improved the reflectivity but not the scatter (89.8% and 1.80%), sign that the mirror was probably not clean enough when coated.

    M2 reflectivity [5] and M2 scattering [6]

Reflectometry and BRDF at the Schmidt: measured in April 01

• 81.3% (470nm), 81.4% (530nm), 83.0% (650nm), 82.6% (800nm
• 10.0% (+20deg), 9.5% (-15deg), 9.3% (-45deg)

Last update: 16 April 01

Reflectrometry at the 1.3-m Telescope

Follow-up November 2009 [7]

1.5-m Telescope reflectrometry data

A/ PRIMARY MIRROR:

3 APRIL 01:

Reflectometry and BRDF data with Iris:

• •  85.3% (470nm), 85.7% (530nm), 83.1% (650nm), 82.5% (880nm)
  •  4.6% (+20deg), 4.3% (-15deg), 2.0% (-45deg)

This corresponds to a 1.1% reflectivity loss per month at 530nm (average over 5 months), about twice higher than usual.

SEPTEMBER and NOVEMBER 00:

Regular washing in September improved reflectivity by 8.5% in average. Reflectometry and BRDF data with Iris (before/after wash respectively):

• •  83.1/92.4% (470nm), 83.7/92.7% (530nm), 80.7/89.1% (650nm), 81.1/89.0% (880nm)
  •  4.8/0.1% (+20deg), 4.3/0.2% (-15deg), 5.1/0.02% (-45deg)

These numbers are actually close (some are even slightly higher!) than what is quoted in the litterature for a new coating. In October, rain leaked through the dome shutter and contaminated M1 (lots of long ugly white stripes across the mirror): in the bad areas (about 20% of the mirror), about 6% loss was measured but it was well recovered with washing in November (only about 1.8% less compared to previous wash). The reflectivity and BRDF improved to:

• •  91.5% (470nm), 91.8% (530nm), 87.5% (650nm), 87.6% (880nm)
  •  0.2% (+20deg), 0.2% (-15deg), 0.02% (-45deg)

MARCH 00:

Mirror washed on the 21st, see the plots of reflectivity and scattering [8] (the numbers after the last aluminizing are also plotted for reference). All data with the Minolta.

SEPTEMBER 98 - MARCH 00: Regular CO2 cleaning

See the plots of reflectivity [9] and scattering [10] for that period (data with the Minolta).

Reflectivity decreases in average by 2.5% per year and scatter increases by 2.0% per year. There seems to be a seasonal trend as the slopes are much higher in 'summer' (October to April) when the site humidity is higher (32-55%), which makes the dust stick more easily onto the glass: summer slope for R and SC is about 0.63% per month, whereas it is only about 0.05% per month in 'winter' (May to September, RH is 12-32%).

Numerical data: in parenthesis (last column) is the number of points sampled on the mirror for each result. Each point on the mirror is sampled 3 times (averaged by the reflectometer). On some dates (like 10Jan98), we also indicate the variations (+/-) in the data. For the first row, replace "before/after CO2 cleaning" by "before/after Al". The case "before Al" means the mirror was CO2 dusted and measured just before realuminizing.

See the efficiency plots of the CO2 cleaning: reflectivity [11] and scattererometry [12]. Average gain per cleaning session is R+0.70% and SC-0.30% in the visible.

2 SEPTEMBER 98:

Aluminizing, see the plot [13] of reflectivity and scattering before and after aluminizing (last one was in Apr 97).

B/ SECONDARY MIRRORS

Aluminized together on August 26, 1999. Measurements with the Minolta.

See the f/13.5 [14] mirror reflectometry and scatterometry.

See the f/7.5 [15] mirror reflectometry and scatterometry.

Last update: 18 April 01

Follow-up November 2009 [16]

[As part of the move to the NOIRLab web site, this page is not to be migrated and its contents have been archived in https://www.ctio.noao.edu/cgi-bin/DocDB/ShowDocument?docid=1404 [17]]

A recent summary of reflectometry data

Measurements of the reflectivity and total integrated scattering of the Blanco primary mirror (M1) and f/8 secondary mirror (M2) are periodically taken and archived to monitor the quality of the telescopes reflecting surfaces. Currently, measurements are taken with an Iris 908RS scattero-reflectometer during mirror washings (every few months) and aluminizations (less often).

Here is the most recent memo by Daniel Holck on the reflectivity data for the Blanco Telescope:

Blanco Reflectance Log: March-November 2011 [18]

Recent measurements (2011) for the Blanco primary mirror M1 show that the reflectivity in the blue (470-530nm) is 92% after aluminization, and slightly lower at redder wavelengths (90% @ 650nm, 88% @ 880nm). Reflectivity tends to decrease ~0.3-0.6% per month afterwards, depending on wavelength. Recent mirror washings appear to increase the blue reflectivity ~0.3-1.0%. The reflectivity of the Blanco f/8 secondary mirror M2 is measured less often, but given its position and baffling, it is much more protected from dust and less variation has been measured. M2 was aluminized in 2001, and recent washings took place in 2007 and 2010. The M2 mirror's reflectivity is typically ~90% in the blue optical (470-530nm) after a wash.

What follows are archival measurements of the Blanco mirrors for the period 1998-2010, and associated documents and figures.

A/ Primary mirror :

AUGUST 00 - NOW: reflectivity [19] and scattering [20]

The measurements were interrupted between November 00 and February 01. Monthly variation rates (over the last 8 months) are: R -0.34% and BRDF(-15) +0.24%. The scatter in the reflectometry plot since February is not well understood, although it might be partly due to the summer seasonal effect (high humidity and thus unstable measuring conditions?).

FEBRUARY 01:

in-situ washing. Iris reflectometry and BRDF data before/after wash was respectively:

* 89.6/93.5% (470nm), 89.6/93.6% (530nm), 86.3/89.8% (650nm), 85.1/89.0% (800nm)
* 2.66/0.21% (+20deg), 2.36/0.33% (-15deg), 0.47/0.24% (-45deg)

AUGUST 00:

Aluminizing. Iris Reflectometry and BRDF data before/after the coating was respectively:

* 70.6/92.6% (470nm), 71.7/93.0% (530nm), 70.1/89.3% (650nm), 72.2/88.0% (800nm)
* 11.4/0.13% (+20deg), 13.1/0.25% (-15deg), 1.02/0.03% (-45deg)

Before the coating process started, we noticed a strong reflectivity discrepancy between the Minolta and Iris measurements, of the order of 10% (Iris: 71.7% vs. Minolta: 81.5% at 530nm), which becomes much less for the fresh Aluminium (Iris: 93.0% vs. Minolta: 93.6%).

AUGUST 98 - AUGUST 00:

Regular CO2 cleaning (twice per month)

All data taken with the Minolta. See the plots corresponding to the entire period: reflectivity [21] and scattering [22].

Yearly variation rates are R -2.5% and SC +2.0%, in agreement with what is measured at the 1.5m. A seasonal trend seems to show up, as is the case for the 1.5m.

Below is a table of the numerical data. In parenthesis (last column) is the number of points sampled on the mirror for each result. Each point on the mirror is sampled 3 times (averaged by the reflectometer). On some dates (like 23Dec98), we also indicate the variations (+/-) in the data. For the first row, replace "before/after CO2 cleaning" by "before/after Al". The case "before Al" means the mirror was CO2 cleaned and measured just before realuminizing.

From 17 April 99 to August 00, we sampled the same 3 points marked on the mirror (the sampling points were random up to that date). On 18 Sept 99, we changed the edge sampling point.

Here are efficiency plots of the CO2 cleaning: reflectivity [23] and scattererometry [24]. Average gain per cleaning session is R+0.36% and SC-0.28% in the visible.

AUGUST 98:

See the plots for before washing [25] and after realuminizing [26] (note: the mirror had not been cleaned at all since early 1997). Data with the Minolta.

2/ SECONDARY MIRRORS:

# F/8 M2: last aluminized on 31 January 01. Iris reflectometry and BRDF data before/after coating was respectively:

* 80.7/90.3% (470nm), 81.6/90.5% (530nm), 80.8/88.2% (650nm), 80.1/85.5% (800nm)
* 6.9/0.2% (+20deg), 6.3/0.2% (-15deg), 0.2/0.03% (-45deg)

Blanco R&S follow up [27]

Blanco Follow up January 2010 [28]

Last update: 23 November 2011 by Eric Mamajek

Instrucciones para el uso del IRIS 908RS

Max Boccas, 20 Agosto 2000

Reflectivity measurements are made with 4 leds at 470, 530, 650 and 880nm at 45deg incident angle (the data is then corrected to 0deg with the Fresnel laws). The scatter is measured through the BRDF (Bidirectional Reflectance Distribution Function) and measured with a laser diode at 670nm (this is sensitive to polarization). The BRDF is the scattered power per unit solid angle normalized by the incident power and the cosine of the scatter angle. The BRDF is measured at 3 angles from the normal (0, 30 and 65deg) and for the incident angle 45deg. The log records BRDF(65,45) noted '+20deg', BRDF(35,45) noted '-15deg' and BRDF(0,45) noted '-45deg'. Micro-roughness of the surface is calculated from the scatter.

1. Poner la maleta del Iris a temperatura ambiente unos 30min antes de la medición
        (dejar maleta abierta para estabilizar más rapido aún).

2. Encender el aparato en ON. Si no muestra 'Low Battery', seguir asi.
            Sino, enchufar el cable del transformador y conectar al 110V (el Iris debe estar ON primero sino se pondrá en modo Carga de bateria).

3. CALIBRAR

3.1. Calibrar la reflectividad

• Amarrar el Reflectivity Gauge al Iris (asegurarse que los pins entran bien).
• F2 para entrar en modo calibracion
• F3 si no está en modo Reflectividad
• START
• Cuando terminó la calibración, aparece una pantalla con un cursor que titilea para pedir una referencia: escribir CAL
• ENTER
• START
• Chequear que los valores entregados (Blue, Green, Red IR) corresponden dentro de 0.2% con los escritos en el Gauge
• ENTER

3.2. Calibrar el Scatter

• ESC
• F2
• F3
• START
• Cuando termina aparece una pantalla con los 3 ángulos: sacar el Reflectivity gauge, guardarlo y poner el Scatter gauge.
• START
• Cuando termino la calibración, aparece de nuevo la pantalla con el cursor que titilea
• ENTER
• START
• Chequear que los valores entregados (+20, -15, -45) corresponden dentro de 0.2% con los escritos en el Gauge
• ENTER
• ESC

4. PRE-MEDICIONES

Limpiar 3 pads del Iris con alcohol y llevar Iris al espejo.

4.1 Scatter

• La pantalla está en modo mediciones de scatter y el cursor pide una referencia: escribir 4M1
• ENTER
• START
• Repetir Enter/Start para tener 5 puntos de mediciones
• ENTER

4.2. Reflectividad

• F2 (modo medición Reflectividad)
• ENTER
• START
• Repetir Enter/Start para tener 5 puntos de mediciones
• ENTER

5. Limpiar espejo con CO2 (dejar Iris protegido)

6. POST-MEDICIONES

• ESC
• La pantalla está en modo mediciones de reflectividad y el cursor pide una referencia: escribir 4M2
• repetir las etapas del item 4.

7. Apagar y guardar el Iris. Tener mucho cuidado con la limpieza de los gauges y el Iris mismo.

8. Descarga de la data al PC. Conectar el Iris a la puerta serial del PC. Lanzar programa Hyperterminal con el icono Iris908rs. Seleccionar Transfer/Capture_Text y el file C:/max/iris/capture.txt

Apoyar en F4 y START en el iris. Aparecerá la data en la ventana del PC. Cuando termina cerrar el Hyperterminal y apagar el Iris.

NOTES ABOUT CALIBRATION:

The Minolta uses a white shiny surface as calibration gauge and is more oriented toward colorimetry analysis. The absolute calibration of our device is wrong (Al reflectivity in the blue has a peculiar increasing slope and sometimes approaches 96%!) so only relative measurements (loss and gain) are meaningful. On the contrary, the Iris uses both a reflection gauge (a polished and protected aluminized mirror) and a scatter gauge (a ground glass) that are precisely calibrated in factory, thus providing absolute numbers that can be directly compared with the theorical curves presented. Unfortunately, our experience trying to calibrate the Minolta versus the Iris is not very coherent as we see different behaviours with different coatings: in general the Minolta reflectometry data seems to be overestimated by about 3.7% (470nm), 3.1% (530nm) and 5.0% (650nm) for coatings made at the 4m chamber, and by 1.3% (470nm), 1.0% (530nm) and 3.1% (650nm) for coatings made at the 1.5m chamber (all offset are within +/-0.5%).

Minolta reflectometer instructions for data downloading to a PC

Minolta CM-2002 spectrophotometer

Characteristics: solid angle of the measurement beam: 7.4deg ; incidence angle of measurement: 8deg ; measured area: diameter 8mm

In mode SCE (Specular Component Excluded), an internal light trap opens up and let the specular reflection of the source lamp onto the sample surface be captured and eliminated from the measurement. Only scattered light from the sample surface is measured. The Total Integrated Scatter (TIS) is measured and relates to the micro-roughness of the surface as: In mode SCI (Specular Component Included), the light trap is closed, therefore the source light bounces off it and gets to the sensor: the combination of reflected and scattered lights are measured. Specular reflected light will then come from the calculation of the difference SCI-SCE.

Procedure for internal memory data transfer to PC; 11/08/98 (M. Boccas)

1. PC and Spectr OFF: connect the RS232 cable to Spectr and COM port of PC

2. Turn ON PC and Spectr.

3. On Spectr:

- press DEL and MENU to go to menu

- press AVG to select DATA OUT

- press the DOWN KEY to enter data out menu

- press AVG to select SETUP and enter with DOWN KEY

- Select 9600 bauds, Even, 7 bits/char, 1 stop bit, Auto out Off, output to PC

- press MENU, press AVG to select REMOTE, enter with DOWN KEY

- the Spectr shows /REMOTE/ and is ready for data transfer

4. On PC (Windows):

-enter Start/Programs/ Accessories/Hyperterminal/Hyperterminal

- give name to your emulator/hyperterminal (for example "reflectometer")

- select the COM port used

- enter the same settings as in the Spectr (9600, even, 7, 1) and NONE for Flow control

- enter the FILE menu, choose Properties, then Settings, then Ascii setup, select "Append line" in the Ascii receiving section (so you can see all the commands and received data on different lines of your emulator)

- in the menu TRANSFER, start the Capture Text option

- Define a capture file (create a dummy one before) or use the default one proposed

NOW

- type DQR then Return to see the number of data sets in the Spectr memory card

- type SDR then Return

- type the number returned by DQR then Return
(see list of commands p86 Minolta manual)

- the Spectr display will show /OUTPUT/ during transfer
(Press DEL on Spectr anytime to interrupt transfer)

- the transfer ends with a E in the hyperterminal window

- stop the Capture in the TRANSFER menu

- put the data in Excell for plotting

Note: the SDR command downloads all the files in the memory card up to the number specified. If you haven't erased the memory card since last measurements, you will need to select only the wanted last files in Excell.

Total Integrated Scatter:

In theory, scattered light dependence on wavelength is given by:

TIS=(4.pi.d/l)^2

where TIS is the total integrated scatter measured in 2pi str, d is the roughness height and l is the wavelength. Therefore, scattered light should be lower in the nearIR than in the blue. Particle size might modify that law if they are about the size of the wavelength.

Reflectivity of Al, Ag and Au. For bare Al coating: 92.1% (470nm), 91.6% (530nm), 90.5% (650nm) and 88.0% (880nm). Data from Hass G., 1965, Applied optics and optical engineering, ed. Kingslake, Academic Press, New York, vol. III, p309

[29]