You are on page 1of 17
Guide to Selecting Standards Laboratory Environments Recommended Practice RP-14 December 2005 i Wire scise NCSL International Guide to Selecting Standards Laboratory Environments Recommended Practice RP-14 December 2005. Prepared by NCSL International ‘ENCSL Intemational 2005 ‘All ghts reserved RP.14 ‘Selecting Standards Laboratory Environments Preface NCSL International (NCSL!) has prepared this Recommended Practice (RP) toward the goal of Uniformity in the field of instrumentation and measurement. To be of real value, this document should be not static, but subject to periodic review. Toward this end, NCSLI welcomes com- ments and suggestions for improvements, which should be addressed to the President of NCSLI ‘at 2995 Wilderness Place, Suite 107, Boulder, Colorado 80301-5404. This RP was initiated and prepared by the Laboratory Facilities Committee and approved for Publication by the Board of Directors in November 1998. The members of the Laboratory Facilities Committee are David W. Braudaway, Co-chair Sandia Laboratories (retired) Doug Cooper, Co-chair TAC Americas Gloria J. Neely Naval Warfare Assessment Center, Measurement Science Directorate Robert E. Williams Naval Warfare Assessment Center, Measurement Science Directorate NCSLI Information Manual it December 2005 Permission to reproduce Permission to make fair use of the material contained in this publication, including the re- production of part or all of its pages, is granted to individual users and nonprofit libraries provided that the following conditions are met: (1) The use is limited and noncommercial in nature, such as for teaching or research purposes. (2) The NCSL international copyright notice appears at the beginning of the publication. (3) The words *NCSLI Information Manual" appear ‘on each page reproduced. (4) The following disclaimer is incorporated andior understood by all persons or organizations reproducing the publication: “Republication or systematic or ‘multiple reproduction of any material in this publication Is permitted only with the written permission of NCSL International.” Requests for such permission to reproduce should be addressed to NCSL Intemational, 2995 Wilderness Place, Suite 107, Boulder, CO 80301-5404, Permission to translate Permission to translate part or all of this publication is granted provided that the following conditions are met: (1) The NCSL Intemational copyright notice appears at the beginning of the transiation. (2) The words "Translated by [enter Translators name)" appear on each page translated, (3) The following disclaimer Is included and/or understood by all persons or ‘organizations translating this publication. If the translation is copyrighted, the translation must carry a copyright notice for both the translation and the publication from which itis translated. Disclaimer ‘The materials and information contained herein are provided and promulgated as an industry aid and guide and are based on standards, formulae, and techniques recognized by NCSLI. The materials are prepared without reference to any specific international, federal, state, or local laws or regulations, and NCSLI neither warrants nor guarantees any specific result when the materials are relied upon, They provide a guide for recommended practices and are not all-inclusive. Printed in the United States of America iv NCSLI Information Manual RP-14 Selecting Standards Laboratory Environments Contents Preface Permission to reproduce Permission to translate. iv Disclaimer ..... iv 1. Purpose.. 2. Introduction and background .. 2.1 Factors Affecting Mission. 3. Temperature 3.1. Selection of nominal temperature... 3.2 Limits on temperature variation. 3.3 Temperature records Vibration... Relative laboratory air pressure Sz }. Humidity and Air Pressure; Correlated Envi Illumination... Particulates... . Electromagnetic interference.. 10. References. SPngegan NCSLI Information Manual RP-14 Selecting Standards Laboratory Environments Guide to Selecting Standards Laboratory Environments 1. Purpose Standards laboratory environments must be carefully selected to meet the specific mission of each laboratory. The mission of the laboratory depends on the specific work supported, the measurement disciplines required, and the level of uncertainty required in the measurements. The purpose of this document is to provide guidance in the selection of laboratory environments suitable for maintenance of | standards and calibration operations. [tis intended not to mandate a specific environment for a specific calibration but to direct selection of the environment and to offer some suggestions on how to extend precision in an existing or achievable (practical) environment. Although this document pertains specifically to standards laboratories, it can be applied to any laboratory requi 2. Introduction and background Environmental conditions in standards laboratories must be maintained within limits sufficiently strict to ‘enable the required measurements to be made. The limits depend on the type and required uncertainty of the calibrations done in the laboratory. The application of new equipment and techniques necessitate a review and, possibly, a revision of requirements, The specific mission in any measurement discipline may dictate stricter or less stringent requirements. The reasons behind this selection are often economic, and the impact necessitates agreement with the customer to achieve an acceptable level of calibration service. Additionally, where the customer needs cannot be met by a reachable mission within the calibration ity, arrangement for outside calibration may be necessary NCSLI Information Manual 1 December 2005, 2.44 Factors Affecting Mission ‘The mission of a laboratory is determined by consideration of several factors: 1. ‘The range of values, the uncertainty expected for instruments and standards to be calibrated, and the customer's requirements; 2. The availability of appropriate standards, calibration equipment and measurement techniques to cover the range of values and meet the uncertainty or uncertainties necessary to mect customer's requirements; 3. Funding necessary to support overhead and personnel time necessary to perform measurements (0 the required level; this cost of calibration is often a compromise between instrument and/or standard specifications and funding available. ‘This guide is based on both experience and results of studies that have been conducted on ways to identify 1 satisfactory environment. Accordingly, it covers both areas where environmental conditions can be relaxed and those where they must be tightened. This document is organized from the most important parameter (or set of parameters) limits to those of less importance. Some proven techniques for circumventing environ-mental limitations are described; the need for assessing the effects ofa recent out- ‘of-control condition is also emphasized. Where practical, requirements are parallel to those of ISA- RPS2.1-1975, Recommended Environments for Standards Laboratories (1}. For additional helpful information, see ISO 554:1976, Standard atmospheres for conditioning and/or testing, (2] and NCSLIRP- 1, Laboratory Design (3] NOTES: 1. Caution is advised in referencing published standards. All standards have a mandatory five-year review with an update or reaffirmation expected. Many standards appear to lie beyond the five-year review; thus, they may contain recommendations that have been changed or are no longer active. Their terminology varies: some prefer to use ‘metrology or calibration rather than standards. In the definitions of these words, standards is the most genera. 2, ISA-RP52.1 isa recommended practice rather than a standard. Itwas written ata time when the general approach to laboratory environments was keyed t0 a position in a relatively rigid hierarchy. Even with strong intent, many have found that they cannot scet the hicrarchy requirements for those disciplines needed to meet their laboratory mission; for some, more stringent environmental constraints are warranted, and for others, the constraints can be relaxed. Accordingly, the preference within NCSLI has been to specify the environmental needs based on evaluation ofthe laboratory mission rather than to ty to fit a hierarchy. Although the approach of this document differs somewhat from that of ISA-RPS2.1 in the selection of environments, ISA-RP52.L contains much information that is still pertinent 2 NCSLI Information Manual RP-14 Selecting Standards Laboratory Environments, 3. NCSLI RP-7 contains detailed information on the steps and requirements that are necessary in upgrading an existing laboratory or designing a new facility. It directs the laboratory staff and design staff to the myriad of standards that govern the various aspects of design and construction, mn looking at the requirements necessary for a specific measurement mission, we need to evaluate the costs and advantages associated with constructing a completely new facility and those associated with ‘modernizing or upgrading an existing facility. A compromise solution is to fit modular laboratory rooms into an existing facility. This approach provides an environmentally controlled space within a controlled cavironment that can lead to reduced demands on the module control system and be economically attractive. 3. Temperature 3.1 Selection of nominal temperature Only two nominal temperature values are used for environmental control of standards laboratories: 20°C and 23 °C. The choice is dictated primarily by prevailing practice in the particular standards discipline: 4. 20°C: dimensional and mass ». 23 °C: force, electrical (do, ac low and high frequency), and most physical disciplines ‘When operation at a temperature other than 20 °C or 23 °C is necessary, an evaluation of the effects ofthe nonstandard reference temperature is recommended. The evaluation should include an analysis of the cor- rections to values, the increase in uncertainty of measurements due to the nonstandard temperature, or both. Documentation of the rationale for the choice of nonstandard temperature is recommended. NOTE: For those measurement disciplines in which the use of dimensional standards is frequent, 20°C may be the preferred choice. However, electronic equipment operated in an ambient temperature different ftom the 23 °C calibration temperature usually has an offset (bias) for which the correction may not be known and may be difficult to determine. The values of properties of artifacts (c.., dimensions of gage blocks and resistances of electrical resistors) that have well established temperature coefficients can be corrected for use at a temperature other than the reference temperature, provided that their actual operating temperature can be established and controlled, NCSLI Information Manual 3 December 2005 Table 1 Laboratory temperature variation limits Variation ‘Typical Application '40.01°C Highest precision dimensional measurement, direct measurement by physical or interferometric methods. ‘The limit of measurability may be computed by multiplying the temperature variation by the expansion coefficient (e.g., (um/m)/°C)] and combining this with the resolution limit of the measuring apparatus, Precision dimensional and mass calibration. In dimensional measurements, direct and comparison measurements to 1.5 un/m for steel or 5 m/m for aluminum. ‘Mass comparison of similar density weights with repeatability (buoyancy corrected): uncertainties of 0.1 mg or better at 1 kg (actual useable values depend on balance resolution and measurement technique). 20.25°C High-quality mass calibration, precision force measurement, and high-precision de calibration, Repeatability (buoyancy corrected) of 0.25 mg or better at | kg for weights with similar densities. Force measurement with a resolution limit of $0 NAN (lbf/Ibf), 40.5°C Optics and general dimensional calibration, Direct and comparison measurement to 5 ynnvim of length for steel or 15 im/m of length for aluminum. 41°C ‘ac, microwave, and most physical disciplines at the highest calibration levels; ‘general de calibration. de calibration at 10 pV/V, 10 4, ete., and less precise levels; with mea- surement of temperature or use of controlled baths, more precise work may be done. 22°C Radiation, general calibration in ac, microwave, and most physical disciplines. Precise figures on ac calibration levels are not available; general practice is stipulated NOTES TO TABLE | ON TEMPERATURE VARIATION LIMITS: 1. Measurement of the actual temperature of the artifact being calibrated is recommended to reduce measurement uncertainty. The temperature of the artifact is more constant than the temperature ofthe laboratory environment owing to the averaging (integrating) effect of equipment and the thermal time constants of the artifact. 4 NCSLI Information Manual RP-14 Selecting Standards Laboratory Environments 2. Temperature may meet the selected variation constraint but not be at nominal value owing to ‘uncontrollable room temperature gradients. Gradients can arise from nonuniform airflow, heat dissipated by the artifact, or measuring equipment, and heat radiated from personnel, Measurement of actual artifact temperature is necessary for meaningful corrections to be applied. 3. Heat sources, including lights and personnel can be especially troublesome in strictly controlled environments. An adaptive control system can be advantageous in compensating forthe heat output of one ‘or more persons (approximately 100 W per person). 4. Extended, nonmeasurement, warm-up times are required for most high-precision balances to achieve stable operation, The principal contributor is heat from the operator. Reduction of nonproductive operator time can be achieved by using a heat source to warm the equipment. For more information on the technique, soe Braudaway, D.W., “Minimization of operator-balance interactive warm-up by use of an alter ego heat source,” IEEE Trans. Instrum. Meas., Vol. IM3S, no. 4, part 1, pp. 404-408, Dec. 1986 [4]. 5. Floor-to-ceiling vertical laminar flow of air is recommended to minimize gradients in environments controlled to +0.01 °C or 40.1 °C. One or more small rooms are necessary to accomplish the result. For other control limits, normal air conditioning practice is adequate. 6. Fordimensional equipment, the time to achieve acceptable equilibrium with the environment s long, ‘owing to machine and artifact thermal time constants. During the equilibration period, measurements are ‘not practical even though the air temperature meets its required limit, unless the temperatures of test and reference artifacts are measured and corrections are applied. 7. Averaging of temperatures from several points in a laboratory can be used to produce a more reptesentative value of the controlled temperature. Electrical sampling of a number of points can provide a readout of each temperature and an average ofthe sampled temperatures for the control input, Some systems, however, operate in an unsatisfactory mannet in which the last point sampled is the control input rather than the average of the sampled points. A tree structure can also be used to provide the average temperature in a room, In this approach, air is <érawn in through arms of the tree and across a single temperature sensor before being retumed to the room, Because the physical time constant of this process is larger, this system is immune to minor local perturbations of room temperature, 3.2 Limits on temperature variation The allowed temperature variation changes with uncertainty level and discipline. In Table 1 above, the six temperature variation limits specified are keyed to their general application, For the mission of each laboratory, the level must be selected to match the requirements. Some methods of circumventing the ‘undesired thermal effects are given in the notes on temperature variation limits below. NNCSLI Information Manual 5 December 2005 SPECIAL NOTE ON VERY TIGHT TEMPERATURE CONTROL: Very tight temperature varation limits (.e,,#0.001 °C to 40.01 °C), as required for standard cells (typical reference temperature of 28 °C or 30 °C) and precision resistors (reference temperature of 25 °C), are normally achieved by use of ‘an oil bath or an air bath. These baths require a laboratory temperature control of only moderate ‘degree against which to achieve their internal temperature values. The laboratory variation is reduced bby a factor of 200 to 10,000 in such baths; however, this reduction factor is not normally specified. It {is usually necessary to monitor the apparent temperature inside the bath to ensure the required degree ‘of control and to correct the value. In application, resistors dissipate power, but the cooling by the fluid of the bath precludes a significant temperature rise 3.3 Temperature records Toknow that the temperature is within acceptable bounds at any time and that it as remained sufficiently s0 for an acceptable period, itis recommended that a continuous permanent recording of the air tempera- ture be maintained for each controlled laboratory area. However, many artifacts (e.g. resistors) and most ‘measuring equipment (electrical and/or electronic or physical and/or dimensional apparatus) di power. In use, the temperatures of these devices are from 0.1 °C to about 10 °C higher than ambient temperature, (Shunt temperatures run much higher than this at rated current.) Precise metrology often requires measurement ofthe actual temperatures of equipment for correction of value(s) and for assurance thatthe temperature is sufficiently constant during measurement. Such actual temperature values are not readily amenable to recording; however, the temperature used in applying corrections must be measured with a calibrated instrument. The following general measurement rules apply: ‘a. Measurement in the physical disciplines and in most of the electrical disciplines is acceptable if the laboratory temperature is in control and has remained so for a suitable period. An arbitrary four-hour period which is the duration of the warm-up specified to stabilize some electronic instruments, i often used, but some instruments and physical artifacts may require an ovemight ‘warm-up. High-precision resistors used in air are an exception: the internally dissipated power ‘mandates both measurement of temperature duting use, and either application of temperature coefficient correction or operation at significantly degraded uncertainty levels. b. Weighing a mass has special requirements that depend on balance design. Two-pan balances do not show large effects from operator heat input. Electronic balances require the normal warm-up expected for electronic equipment and, in addition, may show an effect from thermal input of the ‘operator. Single-pan balances require specific warm-up operation to minimize the effects of thermal input from the operator before meaningful weighing can be accomplished. This warm-up depends on the balance and may be 45 minutes or more. As discussed in note 4 on temperature variation limits (section 3.1), an “alter ego” heat source may reduce warm-up to a more practical time. Itis recommended that the weights be allowed to com to an equilibrium with the environ- ‘ment inside the balance chamber to preclude circulating air currents and the resulting bias forces. Weight-balance equilibration time depends on the mass of the weights and the characteristics of the balance. 6 NCSLI Information Manual RP-14 Selecting Standards Laboratory Environments © Dimensional measurements are affected significantly by the past and the present temperature of the laboratory. In general, laboratory temperature being within variation limits is @ sufficient condition only for relatively low-precision measurements. For precise measurements, the measuring equipment and the reference and subject artifacts must be allowed to come to substantial equilibrium with the laboratory environment, Because measuring equipment has a large thermal mass and low conductive coupling to air, itrequires substantial time to recover from ‘thermal offset. The time constant depends on the equipment, with a range from a few hours to ‘more than a week for large coordinate-measuring machines. Depending on the magnitude of the thermal offset to which machine and artifacts may have been subjected, several time constants under controlled conditions may be required in advance of the measurement. The problem is exacerbated when a low-conductivity material, such as granite, is involved. For such materials, significant temperature gradients and dimensional offsets within the structure are expected for an extended period of time. Evaluation of the recent history of laboratory control and the behavior of artifacts under testis recommended for all measurements. Measurement of the temperature ofthe artifact under test is often helpful and may be necessary to enable a precise correction for the ‘temperature effect, Such a measurement can be used to extend useful machine operating time since the long time constant integrates (averages) over temperature variations, and it may produce « temperature inthe test with less variation than that inthe laboratory environment, Availability of record of the laboratory environment is recommended so that the adequacy of environmental conditions can be ensured prior fo attempting measurement 4. Vibration ‘Vibration is an important factor in measurement. The effets of vibration are most noticeable in light- beam galvanometers (nanovolt amplifiers), mercury columns, mass balances, and dimensional- ‘measuring equipment. The discontinuance of the use of freely suspended galvanometers has sig- nificantly reduced vibration effects in de laboratories because nanovolt amplifiers show a much lower response and electronic detectors show almost no response. Also, for electronic balances, the effects of vibration are significantly less than those for single- or double-pan balances; however, the sensitivity of electronic balances is also lower. Mass balances are the most affected by vibration, bu it is recommended that a low vibration level be maintained throughout the standards laboratory Achicving a low vibration level is a matter of proper building design and site selection. An isolated floor pad is very effective in lowering vibration levels. The recommended vibration control goals are (0.25 um (10 pin) maximum displacement amplitude from 0.1 Hz to 30 Hz and 0.001 g maximum from 30 He to 200 Hz. The equipment itsefis often the best indicator ofa vibration problem. Vibration can bbereduced by isolation through specialized mounts or ai-bag supports. An isolated massive block can also be used to reduce vibration. Air-bag and isolated massive-block systems make maintaining a level condition difficult when mass positions are changed. NCSLI Information Manual 7 December 2005 In the design and selection of a new facility, a significant design effort toward minimizing vibration should be expended before construction. In existing fecilities, isolation techniques are used to minimize the resulting vibration as needed. 5. Relative laboratory air pressure Laboratory air pressure should be maintained at a positive 12.5 Pa to 25 Pa (0.05 inH,0 to 0.1 inH.0). ‘The purpose isto provide an outflow through leakage channels so that pollutants, particularly partctlates, and cold or hot air will not be drawn into the laboratory through leakage channels and on the necessary short-time open-door conditions. 6. Humidity and air pressure; correlated environmental parameters 6.a, Humidity Humidity, barometric pressure and temperature of an environment are correlated through the hygrometric. relationship. For this reason, complete, independent control ofthese three very important parametersis not possible. For most applications, no control is practical for changes in barometric pressure. Althoagh a correlated, multiparameter control system is possible, such systems are not generally available. As a ‘practical compromise, temperature is normally controlled by using relatively short time constants inthe control parameters while humidity is "maintained" within the needed limits through a control process with relatively long time constants. By this means, a quasi-independent control can be achieved that gives, satisfactory maintenance of both temperature and humidity. Inmany geographic areas, the general climate produces humidity of about 40 %, which is sufficient, and ‘no control ofthe parameter is required as such. The reference point in most approximation formulas i also about 40 % relative humidity, and the effects of moderate variation in the parameter are relatively small. A relative humidity of (40 + 5) % is recommended for all general laboratories. This humidity range enables relatively simple environmental control, and is sufficient to minimize electrostatic effects, yet is, in-sufficient to create troublesome electrical leakage paths. Note, however, that most electrical or electronic equipment is unaffected by humidity in the range of 20 % to above 50% and that the range can be broadened for these applications. For mass, force, and dimensional laboratories, the normally recommended relative humidity is (45 +5) % because the reference humidity for air-density correction equations lies in this range. A lower relative humidity level may be preferred where itis desirable to minimize corrosion of steel measuring instru- ‘ments, butitis necessary to precisely measure the relative humidity and correct for its effects on measure ‘ment parameters, A value of (35 +3) % has been used successfully. This lower relative humidity is more difficult to achieve and mandates a noncondensing environmental control (see note below), which is highly desirable where corrosion or performance deterioration of equipment may result. Some mechanical 8 NCSLI Information Manual RP-14 Selecting Standards Laboratory Environments balances are specified for use in a noncondensing environment, bt this environment is more expensive to ‘implement. Accordingly, itis recommended only for the specified areas where it offers advantage. NOTE ON NONCONDENSING ENVIRONMENT: Most air conditioning systems ‘operate by producing saturated chilled air and then either mixing this moisture-laden cold air with heated air or heating the moisture-laden cold air. This mode of operation is efficient and yields about 43 % humidity in a 20 °C environment. However, system dysfunction may deposit a significant amount of moisture on measuring-instrument surfaces, which can result in corrosion or deterioration of function (not a problem at 23, °C). System efficiency can be retained and humidity lowered by using a small, secondary chilling system, which usually requires both an antifreeze solution for operation and ‘operation of the laboratory inside 2 controlled environment. The artificially lowered hhumicity produced in this double system is a noncondensing environment. Since theit extra cost i significant, such systems are not warranted except where a specific need exists. The reference point for wavelength correction and air-density (buoyancy) correction is about 43 % humidity, but proven formulae and the ubiquitous computer of today can readily make allowance for a lower humidity level. 6.b. Barometric Pressure Barometric pressure variation can be a problem in measurements that run for long time. Among these is the high-precision measurement of mass, where air density isa significant contributor to uncertainty. Calculation of air density has sometimes been done on a quasi-eal-time basis to minimize the effect. In Jength measurements by laser, maintaining the light path in vacuum removes the effect of barometric pressure variation. Alternatively, a wavelength compensation system has been used in which the variation of apparent length in a cavity of fixed geometry from temperature, barometric pressure and humidity is used to correct readings of measurements made with other parts ofthe same laser beam in nearby space. Ina few applications, change of air density or humidity over several hours or days produces effects that cannot be corrected by available processes. The variation of barometric pressure isthe principal problem. Solutions to preclude such variation range from operating sealed, pressure-controlled humidity apparatus to a sealed, pressure-controlled laboratory. However, designing a sealed laboratory is a difficult task, involving consideration of pressure differentials on the structure and designing a complex air-lock system to permit entrance/exit. For such requirements, operation of sealed humicity/temperature/baromettic- pressure chambers is simpler both to achieve and to operate. Accordingly, choice of sealed chambers is recommended. NCSLI Information Manual 9 December 2005 7. jumination Mlumination in laboratory areas should be approximately 1080 Ix (100 Im/t) withall lights on, Individual light controls are required for areas where radiant energy can affect measurements (a low illumination pattern must be established for each specific measurement) and are recommended forall laboratory areas. The heat output of lighting systems is significant. Accordingly, leaving lights on at all times produces @ ‘more constant temperature and is warranted for those laboratories where tight temperature contol is required, 8. Particulates Control of particulate levels in the laboratory atmosphere is recommended but only toa reasonable degree. Limiting particulates in the atmosphere reduces contamination of open oil baths and dimensional-measur- ing surfaces. Direct effects on the measurement require a higher degree of particulate control inthe mass. However, clean-room conditions and specialized clothing are not necessary for standards laboratory ‘operation. Attention to housekeeping is a must; in the most critical areas a tack-sheet to remove shoe- ‘borne particles can be helpful. Filtration of laboratory air by normal means is sufficient in nearly all cases. The wet scrubber used to saturate chilled airs also highly efficient in removing particulates. A potential problem to keep in mind is the slow accumulation of particulates in ductwork and their release at later time. 9. Electromagnetic interference Most precision electrical and electronic equipment s subject to electromagnetic interference (EMD, but to only amoderate degree. Any EMI may be insidious, particularly because detection at microvolt and lower voltage levels can cause bias by partial rectification of the interfering signals. Care should be taken in site selection and use of shielding; the goal isto achieve less than 100 :V/m field strength and 100 wV open- circuit voltage from conducted electromagnetic energy. Neither ofthese specifications is highly restrictive; 100 Vm is about the sensitivity of small AM radio. Equipment in electronics often has # relatively high input impedance that enables open unshielded leads to pick up enough signal to cause problems; thus care in lead dress and shielding is always recommended. For measurement disciplines that have a high susceptibility to EMI, shielded rooms with appropriate attenuation should be considered, Shielded rooms may also be needed where radiation of generated signals. must be suppressed. In addition to possible magnetic pickup, ground loops between instruments and circulating currents from the several connected instruments should be avoided. 10 NCSLI Information Manual RP-14 Selecting Standards Laboratory Environments 10. References 1, ISA-RP52.1-1975, Recommended Environments for Standards Laboratories, ISA (International), Research Triangle Park, NC 1975. 2, 180 554:1976, Standard atmospheres for conditioning and/or testing, Intemational Organization for Standardization, Geneva 1976. 3. NCSLIRP-7, Laboratory Design, NCSL International, Boulder, CO 2000. 4, Braudaway, D.W., Minimization of operator-balance interactive warm-up by use of an alter ego heat source, IEEE Trans. Instrum. Meas., Vol. IM35, n0.4, part 1, pp. 404-408, Dec. 1986. NCSLI Information Manual " Prepared by the Laboratory Facilities Committee First Edition — 1999 ‘Second Edition — December 2005 NCSL International 2998 Wilderness Place, Sulto 107 Boulder, Colorado 80301-5404 (203) 440-3339 ISBN 1-58464-023-5