Thermalspray part count dilemma

Sponsored Links

This article is directed towards thermal spray companies that use travelers in their operations and have operator sign off for each operation performed. Generally thermal spray shops that deal with aerospace or medical hardware definitely use an operator sign off traveler. There are other examples also of quality minded companies in the thermal spray business that use such operator sign off after each operation. As you may know, with an operator sign off traveler, at the end of each operation performed in the entire chain of operations that form the completion of the contracted thermal spray task, the operator responsible for the specific operation is required to sign off along with the part count at the end of his or her specific operation. From a thermal spray quality standpoint, this tells the “world” that the specific signed off operation was performed by the operator in accordance with the written instructions and that the operation was performed on the number of pieces denoted by the operator. While there are other permutations that are possible, in the current treatise we will limit ourselves to discrepancies dealing only with part counts.

An easy way to understand potential problems is by taking a set of practical real world examples. Let us assume that you are in the aerospace thermal spray industry and you have written, approved operation sheets for all of the thermal spray jobs that you receive and process. Let us assume also that a new order arrived. The incoming inspector inspects that parts and signs off stating that he received twenty three parts and logs their dimensions and finds no damage of any kind. The parts then move to cleaning and de greasing. The operator at the cleaning station cleans the parts and signs off stating that twenty three parts got cleaned and then the masking operator signs off twenty three parts received proper masking. So far, things are working perfectly. Next when the parts get to grit blasting, the grit blast operator does his operation then signs off stating he accepts twenty two parts, but incorrectly processed one piece that he tags and sets aside to return to material review board. The plasma spray operator who is next in line signs off twenty two parts and thence to final inspection and then the quality control personnel sign off the certification and the customer gets twenty two parts with a rejected piece that needs to be dispositioned since the error occurred at the grit blast step. While no one wants to make bad parts, a genuine error can occur and as long as the quality checks were properly followed as in our example above, everything is fine.

Now, take the same example above, but let us add a twist to it. Let us introduce an additional operation between plasma spray and final inspection. Let us do a fluorescent penetrant testing for cracks after plasma spray. Now the twenty two parts that were signed off by the plasma spray operator, goes to the FPI operator, who signs off twenty one parts ( due to incorrect part count ) and then the final inspector signs off twenty two parts because he does physically count twenty two parts. Now we have a serious problem. Even though the final inspector is correct as far as part count, there is no guarantee that the FPI operator really tested all twenty two parts and erroneously signed off as twentyone. May be he tested twenty two, one was rejected, but he “forgot” to send it to MRB and let it float along with the other good parts, or may be that there were really only twentyone parts when he received from the plasma person who later discovered that there is an extra part in his booth and then rushed it to final inspect instead of taking it to FPI and upon finding a bin full of parts waiting at final inspect, he decided to put his part also in the same bin along with the rest of the pieces. This is a serious quality problem that has developed now. If I was the customer, I have no guarantee now of the integrity of the operator sign offs. Hence this is a dilemma that needs to get addressed immediately.

Here is a solution to this problem. The quality control procedures must be altered to address these situations as follows: If an operator finds that his part count at his operation, after adjusting for MRBs, rejects, etc, does not match the previous operation’s part count, then the part cannot move to the subsequent operation until the issue of part counts is resolved. This means that if you properly process all the parts that you physically received and no rejects took place and you find that your part count does not match the part count from the previous operation then you cannot sign off your operation, until the issue of the part count discrepancy from two successive operations has been resolved. In the example case above, when the FPI operation accounted for twenty one parts instead of the twenty two that the plasma spray operation accounted for, then the train should have been stopped, the situation analyzed and only then should the FPI operation been signed off and the parts moved over to final inspection.

Seems like a trivial point, but as Michael Angelo said, “trifles make perfection, but perfection is no trifle.”