Are there a standard set of environmental conditions, e.g., temperature and humidity, that must be met during the calibration of standard measuring instruments [calipers, micrometers, etc.? A prospective supplier of calibration services is offering to perform such calibrations on site [to save transportation time and $] but I am concerned that the proposed location does not meet the required [?]controls. :confused:

Standard conditions for almost all dimensional calibrations is 20°C (68°F). Now is where it gets tricky. To hold the uncertainties that you are looking for, temperature range varies.

In a nutshell, gauge blocks and instruments grow with temperature increases. They grow (assuming steel gauge blocks and calipers/mic spindles) at a rate of 11.5µm/m/°C (6.4µin/in/°F) ± about 10%.

Therefore, with each °C change from 20°C, we add to uncertainty of 2.3µm/m (or 0.06µm per 25mm). Uncertainty on a micrometer at ±2.5µm (0.0001") over a 25mm (1") span would dictate that you would not want to stray from 20°C by more than about 5°C (68°F ± 9°F). If your micrometers are less accurate, then you could stray more. If they are more accurate, then you need a tighter environment still. This is all reliant on the base uncertainty of the company calibrating your tools, because if they start high, it lessens the room for error allowable for temperature.

The biggest worry is usually rate of change, because the mass of a gauge block versus the mass of a set of calipers or micrometer spindle is a big difference. If your environment can't stay within about 2°C/hour, then calibrations start really giving false measurements. If the temperature of the instrument and the temperature of the standard vary by much, errors really start adding up quickly (hence the little chunks of plastic on the micrometer are known as 'heat shields', and are meant to slow the transfer of heat from your hand to the micrometer).

Calipers are more forgiving (because of the 10:1 accuracy of a mic vs. a caliper), and you can generally get away with about 20°C ± a ballpark.

I omitted a lot of calculations here, for brevity and sanity that some of the savvy folks reading this should pick up on, such as distributions, etc., but this is a pretty good rule of thumb for a quick environmental check.

Humidity does not effect caliper or micrometer calibration, although as a supplier, I never liked taking my blocks into an environment with >60% humidity because they rust so quickly.

Hope this helps,

Ryan

I can't provide as detailed and elaborate response as Ryan has done.

The two websites I can suggest are:

http://www.a2la.org/
http://www.ukas.com/
Both these have got lots of pdf documentation on various aspects of calibration lab requirements. May be you'll find something there also.
-Atul.

Humidity does not effect caliper or micrometer calibration, although as a supplier, I never liked taking my blocks into an environment with >60% humidity because they rust so quickly.


The lab which does our gage blocks and weights does so in an environment with humidity < 45%. Temperature range is
68.0 +/-1C.

Originally posted by Ken K



The lab which does our gage blocks and weights does so in an environment with humidity < 45%. Temperature range is
68.0 +/-1C.

That is actually a fairly broad temp range for a lab doing gage blocks. Even at 20°C ± 0.25°C, temperature accounted for over 50% of our uncertainty on gage blocks, even with thermal compensation. We also had problems with humidity (we spec'd 25-45% RH, and in the winter it always tried to go below 25%, which is not healthy for electronic devices such as gage block comparators).

I now rarely touch dimensional, and I work in a comfy 23° lab. Life is good.

Ryan

Hello,

I would like to know what is the ideal environment in calibrating dimensional and electrical equipment. Ours is currently 21°C +/- 5°C and 50% +/- 10%. Now, I need to justify this figures. Where did we get these figures? What possible justifications can we give?

Thanks in advance!

cheers,
Lord Ituralde

I would like to know what is the ideal environment in calibrating dimensional and electrical equipment. Ours is currently 21°C +/- 5°C and 50% +/- 10%. Now, I need to justify this figures. Where did we get these figures? What possible justifications can we give?
Here are some figures from a couple of readily available documents --
NCSL Recommended Practice RP-14 (March 1999) Guide to selecting standards-laboratory environments. (NCSL International, Boulder CO USA (http://www.ncsli.org/)) ISA Recommended Practice ISA-RP52.1-1975 (June 1975) Recommended environments for standards laboratories. (ISA, Research TrianglePark NC USA (http://www.isa.org/))

TEMPERATUREDimensional, Optical and MassNSCL: 20 °C +/- 0.5 °C for general calibrations

ISA: 20 °C +/- 1 °C overall and +/- 0.3 °C at the point of measurement

All other disciplinesNCSL: 23 °C +/- 2 °C for general calibrations

ISA: 23 °C +/- 1.5 °C

RELATIVE HUMIDITYDimensional, Optical, MassNCSL: 40% +/- 5% RH at 20 °C

ISA: 45% RH maximum at 20 °C

All other disciplinesNCSL: 40% +/- 5% RH at 23 °C

ISA: 20 - 55% RH at 23 °C

Each of these recommended practices contain a lot of other information, of course. Both of them are targeted at "standards" laboratories; if you are in a company's production lab, this is the type of lab you would send your transfer standards to. In both cases, the values above are the most relaxed but still might be tighter than what you want to maintain. On the other hand, they also agree with what I remember (10 years ago) of the US Navy's requirements for general calibration labs -- Measurements using gage blocks: 20 °C +/- 0.5 °C and 10 - 45% RH

Other dimensional/optical/mass: 23 °C +/- 3 °C and 20 - 60% RH
Electrical/Electronic/RF/Microwave: 23 °C +/- 5 °C and 20 - 60% RH.

What you have appears to be a compromise. Do you perform all calibrations in the same aera? The temperature range means that you are giving up some accuracy capability in dimensional measurements, and the high humidity is inreasing the risk of corrosion (rust) on your standards. The only way to get better control is to divide the lab into two areas with sepatrate temperature/humidity control systems. (There may be other ways, maybe an air-conditioning engineer would know ...)


You might also want to check the NIST (http://www.nist.gov/) web site. A couple of items that may be useful are the State Weights and Measures Laboratory Handbook (NIST Handbook 143), and the Gage Block Handbook (NIST Monograph 180).

Reading this thread brought to mind a few questions:

1)If you calibrate all of your gages in a laboratory with a controlled environment, then use the gages in an uncontrolled environment, what kind of results do you get?

2) Does it make sense to let the standards "soak" and calibrate the equipment in the environment it will be used in?

3) Is it better to keep gages in the environment they will be used in, or in a controlled lab?

Reading this thread brought to mind a few questions:

1)If you calibrate all of your gages in a laboratory with a controlled environment, then use the gages in an uncontrolled environment, what kind of results do you get?

2) Does it make sense to let the standards "soak" and calibrate the equipment in the environment it will be used in?

3) Is it better to keep gages in the environment they will be used in, or in a controlled lab?

I think it depends onthe accuracy you need, but some folks surely need to think about it. People making very precise measurements need to maybe do a few tests and calculations and look at the data and decide. The standards labs have to be concerned with much more accuracy than I do for my mics and calipers. If I consider a 1" gage block to have a 15 ppm/C. thermal expansion coefficient and the temp is either 15 or 25 degrees C (59-77 F)the block will change size by ~ .000075" from its size at 20 C. I never calibrate at those extremes and this is more accuracy than I need anyway. In the shop area, let's say the A/C is dead one day and it gets to 32 C. (90 F). My gage block is now ~ .00018" oversize -- still very unlikely this would cause me any measurement issues, but I would not cal at that temperature. If the tool (mics or calipers) is hot any effect is cancelled-out by zeroing it at the use temp. Other folks may need more accuracy so they would need to look at their particular situation.

As for a electronic equipment, same deal, do a few tests and calculations and look at the data and decide.

JMO.

Hello,

I would like to know what is the ideal environment in calibrating dimensional and electrical equipment. Ours is currently 21°C +/- 5°C and 50% +/- 10%. Now, I need to justify this figures. Where did we get these figures? What possible justifications can we give?

Thanks in advance!

cheers,
Lord Ituralde

I seem to remember that you had a Datron/Wavetek/Fluke/whatever 1281. The accuracy spec for that particular piece of equipment is 21 - 28°C, which may pose a problem for you with your present environmental specification.

Many labs, including ours, have separate rooms for dimensional and electrical calibrations, with entirely different specifications. The crux of the problem is that 23°C is standard laboratory temperature for electrical, and temperature does have measureable effect on high-accuracy equipment. Dimensional, on the other hand, has a standard laboratory temperature of 20°C, and variations from that temperature will have a measureable effect with very little temperature change.

You also may wish to worry about your humidity specification, as 60% RH on tool steel tends to cause rust.


Ryan

Reading this thread brought to mind a few questions:

1)If you calibrate all of your gages in a laboratory with a controlled environment, then use the gages in an uncontrolled environment, what kind of results do you get?

No environment should be uncontrolled. The gages must be calibrated in a controlled environment to give you a basis to measure/calculate errors in other environments, such as the shop floor, etc.

2) Does it make sense to let the standards "soak" and calibrate the equipment in the environment it will be used in?

It does not generally make sense to "soak" standards and calibrate in a non-standard environment due to different thermal expansion properties. Basically, what is the material(s) that comprise the item being calibrated, what is the material(s) that comprise the standard, and what is the difference between their thermal expansion? Example: Gage blocks can be made of tool steel, chromium carbide, tungsten carbide, or ceramic. Most tools, such as micrometers or calipers, are made of entirely different types of steel, such as stainless. You now have a difference in thermal expansion to deal with that will be very small at or near 20°C, but will be measureable at 30°C, to the point that at 10", it may mean the difference between in and out of tolerance.

3) Is it better to keep gages in the environment they will be used in, or in a controlled lab?

That's a tricky question. If I have to use a micrometer on product that is in a 35°C room, I would want my micrometer to be at 35°C rather than 20°C. The reason would be that the difference in thermal expansion would be much less at that point, and although it's not perfect, it is a more accurate measurement. If my micrometer was calibrated at or near 20°C, and I know it to be accurate, it is my best bet to not have false accept/rejects.

Ryan

1)If you calibrate all of your gages in a laboratory with a controlled environment, then use the gages in an uncontrolled environment, what kind of results do you get?

Every place I have worked, the rule has been that everything portable is sent to the cal lab for calibration. No exceptions. If it's not portable, then an in-place calibration is done.

I don't know how dimensional tools will react to "uncontrolled" environments -- I'm an electronics guy.

As for electronic equipment, operation in the "specified" temperature-humidity-altitude range is normally covered in the specification sheets. However the user may be expected to figure temperature coefficients ... it should be evaluated just in case the environment you actually use it in is not the same as the one the manufacturer thinks it should be used in.

For example: when I was working in a Naval Shipyard, I saw many cases where test equipment that had a "for laboratory use only" label from the manufacturer was in used out on the waterfront and on board ships that were under repair. (Note that when a ship is being repaired, the air conditioning or heating is often turned off!) But if the test procedure calls for using a network analyzer to check an installed radar feed and antenna, you have to take the test equipment to the ship because you can't take the radar mast to the lab.

I seem to remember that you had a Datron/Wavetek/Fluke/whatever 1281. The accuracy spec for that particular piece of equipment is 21 - 28°C, which may pose a problem for you with your present environmental specification.

Many labs, including ours, have separate rooms for dimensional and electrical calibrations, with entirely different specifications. The crux of the problem is that 23°C is standard laboratory temperature for electrical, and temperature does have measureable effect on high-accuracy equipment. Dimensional, on the other hand, has a standard laboratory temperature of 20°C, and variations from that temperature will have a measureable effect with very little temperature change.

You also may wish to worry about your humidity specification, as 60% RH on tool steel tends to cause rust.


Ryan

Hello,

Yes, we do have that equipment.

I thought that at lower temperatures, the measurements in electrical equipment are more stable. Am I wrong?

Thanks in advance.

cheers!

Yes, we do have that equipment.
I thought that at lower temperatures, the measurements in electrical equipment are more stable. Am I wrong?
Thanks in advance.

In principle, electronic equipment can be designed to work over as wide a temperature range as you like. Test equipment that is intended to be used that way is designed for it -- for example, equipment that an airline would use to perform tests on an airplane. The equipment has to work as expected, whether it's in Moscow in the winter or Phoenix in the summer. However, that type of equipment does not have the same type of performance as a calibration laboratory standard. Generally the tradeoff is wider enviromental conditions and wider performance specifications, or laboratrory-grade performance and a very small range of allowable environmental conditions.

Equipment like your Wavetek 1281 - and, in general, most "laboratory" electronic test and measurement equipment - is designed for best operation at a stable temperature in a narrow range. For example, look at the headings and footnotes of the 1281 specification sheet:
The preferred temperature range is 23 ± 1 °C. This is for the 24-hour or the 12-month specification.
23 ± 5 °C is acceptable provided the Selfcal routine is performed any time the temperature changes more than 1 °C from what it was at the time of the last Selfcal.
It "can" be used from 13 °C to 33 °C provided the Selfcal is performed first at 23 ± 1 °C, and the Temperature Coefficient values are added to the measurement uncertainty. (Note that this also implies that you can't turn it off between the 23 °C Selfcal and the time is is used in the "other" temperature.)
As you can see, this piece of equipment really expects a stable temperature (variation less than ± 1 °C); and a temperature close to the standard 23 °C. This is also an example of why it is beneficial to continuously monitor and record the laboratory temperature.

Would it be correct for me to assume that all of your calibration facility is in the same place -- that is, you don't have separate rooms with individual air-conditioning systems? If separate facilities are not possible you have to compromise, which might be what the specification you originally quoted was trying to do. The compromise means tolerating greater uncertainty from the 1281 (because it is too cool) and dimensional offsets from your gage blocks (because they are too warm).

As Ryan said as well, the relative humidity you mentioned (50 ± 10 %RH) is really high for a dimensional calibration lab, and I would be uncomfortable with that in an electronic lab as well. Moisture causes corrosion, creates leakage paths for voltage and current, and promotes mildew and fungus growth. A "do not exceed" limit of 45-50% would be better if your environmental control system can do it. On the other end, it should not go below 20% or you get problems with static electricity, which can damage electronic equipment. I don't know what the climate is like in Arad Doman (I prefer Manticore, myself) but often in the winter it is necessary to add a lot of moisture to keep the humidity above 20%.

Does this help, or have I simply added confusion?