Developing In-House Calibration Capacities

[Bob the QE]

I am in the investigating stages of looking at setting up a calibration program internal to our company. Currently we outsource the calibration of such instruments as calipers, indicators, mics, ring and thread gages and gage blocks. I would like to set up an internal program where we have our top level measurment device calibrated by a certified body (mic standards, master blocks and rings) and then we would do the rest in house. Just to throw numbers out for costing we are talking about 1800 measurement devices and 1700 hard gages (pin, block and thread gages). Thoughts of all kind welcomed.

Thanks in advance.

[Stijloor]

Originally Posted by Bob the QE

I am in the investigating stages of looking at setting up a calibration program internal to our company. Currently we outsource the calibration of such instruments as calipers, indicators, mics, ring and thread gages and gage blocks. I would like to set up an internal program where we have our top level measurement device calibrated by a certified body (mic standards, master blocks and rings) and then we would do the rest in house. Just to throw numbers out for costing we are talking about 1800 measurement devices and 1700 hard gages (pin, block and thread gages). Thoughts of all kind welcomed.

Thanks in advance.

Bob,

Are they all active? Begin with the active gages first. Build from there.

Stijloor.

[arios]

Nice project Bob! Congratulations

I have seen other folks moving on the same direction with success so I wish you the best also. Based on what you are saying I think you are on the right path. Would recomend you only to develop clear work instructions, and a good training program. Since this is a change on your QMS do not forget to consider this newly acquired responsibility into your internal audit program and part of the changes to discuss in your management review meeting.

PS Do not discard the possibility of using occasionally external services for in house calibrated equipment because on certain circumstances they could be your plan B, e.g. if the workload exceeds your capacity.

[Hershal]

With that many devices you can likely make a good business case for internal cal.

You will need first a controlled environment, and for the devices you listed that should be 20 C plus/minus 2 C, and 40-50% RH.

You will need procedures for each kind of calibration, and the people need to be trained in them.

You must learn to calculate uncertainty to assure traceability, as it is one of the two components of traceability; the other is an unbroken chain of comparisons to SI, through National standards, which in the U.S. is typically NIST.

You will need the right equipment, and to calibrate gage blocks you will likely need a gage block comparator if you do not already have it, and same for a super-micrometer to do rings, plugs, and pins. Not sure of the prices, but you can likely get both for around $30K, maybe less for cheap stuff.

One other point, as has been mentioned, is to keep the option open to use accredited calibration services to maintain capacity. Another potential is to have the lab that you want to set up contract-staffed by an accredited lab. Otherwise, I would strongly suggest finding a good calibration tech with a dimensional/mechanical background and hiring him/her.

Hope this helps.

[Bob the QE]

Thanks Hershal and arios

Good information and food for thought if there are any other thoughts or suggestions I am still looking.

Thanks again covers

[BradM]

Bob,

The others guys covered everything pretty well. Apologies in advance if I kind of repeat some of their points.

Don't underestimate the cost of performing calibrations in-house. It's easy to calculate the cost of standards, lab areas and such. But don't forget about the periodic costs of annual calibration of standards, documentation, some level of internal auditing for your program, training, new standards and most importantly, labor. Also, you may want to give some brain time to calibration failures and remedial action if one of your standards is out of calibration. Depending on your risk and all, you might need to have dual standards and perform more frequent in-lab cross verification of your standards.

Like many programs, the labor will be there in the beginning. But as time moves on and say, production is short-handed, which one do you think will take priority? The labor and training must be dedicated, and management must understand that will be there month after month, year after year.

As stated, do not let your documentation slide. It's so easy to fall into a trap of 'getting something calibrated, and I'll do the paperwork later'.

Just make sure that management is ready to commit the necessary resources. If they flinch, I would keep it external. I bet you can purchase data, and start extending the calibration cycles to save some money.

Also, keep on the table a hybrid of some checks in-house, and some being performed by the external calibration source.