Componentes del Control Interno

When the theodolite vertical axis is not aligned with the centre of the tribrach axis, a default exists. It is an assembly error and cannot be compensated. It is only important if a forced centring system is being used.

3.2

DCB, HCC and VCC selection criteria

Instruments designed to calibrate horizontal RTSs and LTs must fulfil a certain number of criteria. Foremost among these criteria is traceability. Although not formally a prerequisite, it is highly desirable that the instrument have a smaller uncertainty than the instrument being

calibrated. Available LTs and RTSs have manufacturers stated angle uncertainties (k =1) in

the order of 2 arc-seconds and 0.5 arc-seconds respectively. Therefore, ideally the expanded

(k=2) uncertainty should be (considerably) less than 0.5 arc-seconds. Precise (e.g. 1 arc-

second) positioning is not essential, but the angle standards must be capable of setting its position and measuring over the full 360º circle. Finally, the specifications set out by

Ingensand [34] in section 2.4.1 remain valid; in particular, the possibility of an automatic test sequence, calibration of the instrument over its full measuring range, suitability for various RTS and LT models, and the qualification of both vertical and horizontal angle reading systems are essential attributes to fulfil.

Recall there are three main components in these instruments; the distance meter (EDM, ADM, and/or IFM), the horizontal circle and the vertical circle. Other parts such as the compensator are calibrated with these components. Ideally we might want a standard that could measure these three components simultaneously to produce a 3D error map. Practically this is difficult, if not impossible. To physically investigate all possible angles and distances that could be measured naturally (i.e. horizontal angles over ±180 degrees; vertical angles over ±45 degrees; and distances up to 50m), one would require a type of igloo shaped laboratory 100 m in diameter and 70 in height. Everything involved with this laboratory

would certainly be colossal; not least the structural, the air conditioning and the cost considerations. Clearly some compromise must be made.

In the present work, the decision has been made to separate their calibration. The EDMs, ADMs, and/or IFMs are calibrated on the 50 m long ESRF DCB. The horizontal circles are calibrated with the HCC and the vertical circles are calibrated using the VCC. All three of these standards are traceable to the SI unit, the metre. Apart from this, the selection of the standards and techniques described in the following are dictated mainly by instrument constraints (e.g. minimum measurement distance), available resources (e.g. laboratory space, instrumentation etc…) and financial constraints.

3.3

EDM, ADM and IFM distance meter calibration

Figure 3.3 Schematic of the ESRF calibration bench. Zoom a) is the instrument station; zoom b) the servo carriage with the instrument and interferometer reflectors and zoom c) the interferometer station. After the zero error has been determined the servo carriage is moved in 10 cm intervals from 2 m to 50 m to determine the instrument cyclic (bias) error. (Drawing prepared from Solid Works design drawings made by B. Perret)

The ESRF has a 50 m long calibration bench (refer to Figure 3.3). An interferometer is installed on a fixed pillar at one end of the bench and the instrument to be measured (RTS or LT) is installed on a fixed pillar or heavy tripod at the other end. The interferometer and

instrument reflector are installed on a servo-controlled carriage. The bench is equipped with an accredited meteorological station which measures temperature, pressure and humidity. Additional temperature sensors are installed at regular intervals along the length of the bench to improve corrections for the variations in refraction along the line of sight.

Figure 3.4 Typical Interferometer (IFM), Absolute Distance Meter (ADM) and EDM distance error curves. Different curves are from different instruments. The IFM and ADM curves are from three different instrument manufacturers. The EDM curves are three instruments of the

same manufacturer and type. The expanded uncertainties (k=2) in these calibration curves

are 0.050 mm for the IFM and ADM curves and 0.165 mm for the EDM curves.

A typical calibration comprises the determination of the zero (index) error, followed by the determination of the cyclic (bias) error by comparison of measured distance displacements with the accredited laser interferometer. Distances are measured by moving the servo-carriage at regular 10cm intervals along the bench. This results in a calibration curve - an example of which is shown in Figure 3.4. This curve can be modelled and applied to considerably improve field measurements. At the ESRF, the enlarged uncertainty for these calibrations is 0.17 mm for a 2m to 50m EDM calibration with an instrument resolution of 0.1 mm; and 0.05 mm for a 0.3 to 50 m LT ADM or IFM calibration.[4, 44]

3.4

Horizontal angle calibration

Horizontal angles are calibrated against the horizontal circle comparator or HCC. The HCC is composed of a reference plateau, a rotation table, and an angle acquisition system (refer to

Figure 3.5). The conception and choice of composite parts of the HCC was made by the author, although several colleagues contributed to design detailing.

The reference plateau is fixed on the rotation table and rotates with it. The heart of the HCC is the angle standard. This standard, the angle acquisition system, or linked encoders system (LEC) is incorporated into the rotation stage. Each of these systems will now be discussed in turn.

Figure 3.5 Schematic of the HCC assembly showing reference plateau e), the rotation table f) and the LEC system a), b), c), and d).

3.4.1

The Reference plateau

The reference plateau e) in Figure 3.5 is a stainless steel 500 mm diameter 40 mm thick plate. It has a regular grid of M6 threaded holes for fixation. There is a 20 mm bore in its centre. This bore is used for the forced centring of instruments installed on the plateau. Finally there is a high precision machined surface 20 mm wide around the circumference; both on the side and on the top edges. These surfaces, shown as g) in Figure 3.5 above act as targets for the capacitive probes used in the determination of the plateau wobble, or inclination, and eccentricity movements.

At the origin, the primary reason a 500 mm diameter was chosen was to permit the

installation of a hydrostatic levelling system (HLS)8_{[45, 46]. Although provision was made}

for the system, it has as yet, not been used. The idea was to be able to measure both the inclination of the plateau and its movement with respect to an exterior reference point. This would provide a high precision standard fully referenced to the gravity system. The 500 mm diameter base line provided is sufficiently large this type of work. It is the author’s intention to ultimately activate this system.

3.4.2

Rotation stage

The plateau and angle acquisition system are installed on a Micro-Controle RV350CC high- performance precision rotation stage (f in Figure 3.5). The stage has the following

manufacturer’s quoted performances: minimum incremental motion 0.001º; absolute

positional accuracy 0.005º; wobble 16 µrad; eccentricity error 4 µm; maximum rotation speed 80º per second. An important choice in the employment of this particular rotation stage is its normal centred load capacity of 6500 N. This ensures that any LT or RTS instrument commercially available on the market can easily be accommodated.

3.4.3

Angle Acquisition System - Linked Encoders Configuration (LEC)

Although commercially available encoders can provide uncertainties in the order of 0.5 arc

seconds (k=1)9_{, the target of the LEC system is to provide a standard with an expanded}

uncertainty (i.e.k=2) in the order of 0.1 arc seconds or better. In principle, the LEC is

capable of reducing residual error in the already excellent commercially available angle encoders to negligible levels.

The LEC uses two Heidenhain RON 905 angle encoders mounted in juxtaposition. One RON 905 is fixed to the main support assembly and does not move. The second RON 905 body is fixed to the main plateau and rotates with it. The two RON 905 encoders are linked

8 _{The author has written several papers on the HLS installation at the European Synchrotron Radiation}

Facility. The first reference provides an overview of the ESRF systems and references to earlier works concerning them. The second reference provides a thorough theoretical and practical review of HLSs.

9 _{See for example the Renishaw RESR angle encoder system with a manufacturers stated uncertainty of}

0.5 arc seconds (http://www.renishaw.com/); or the Heidenhain RON 905, actually used in this work, with a manufacturers stated uncertainty of 0.4 arc seconds (http://www.heidenhain.com); both accessed November 2008.

through a precision alignment shaft assembly. The shaft and encoders are rotated continuously by Micro-Controle RV120CC high-performance precision rotation stage (c in Figure 3.5). Although the maximum speed of the RV120CC is 80 degrees per second, it is typically operated at 20 degrees per second, thus taking 18 seconds to rotate through 360 degrees. The two RON 905 encoder positions are read out simultaneously and continuously. This assembly, referred to as the LEC for the linked encoders configuration can reduce considerably, or even theoretically eliminate altogether, the residual RON 905 encoder errors. [47, 48]