Narrative:

I work for air carrier as a mechanic in their APU shop in ZZZ, us. On jul/fri/00 I was on a post shift of 4 hours of overtime. I was assigned to assemble and test atomizer assemblies. They are allied signal assembly part #36-1193-8, that are used on burner can assemblies that are installed on allied signal APU's GTCP85-129C, -129E, -129H models. Air carrier has a work card that covers the assembly and test of the atomizer assembly, work card #wxxx. I have assembled atomizer approximately 3 other times, but this was the first time I assembled this model. The housings, approximately 11 each, were now out of stock. The part in question, the flow dividers (allied signal part #3883998-1) were also new from the company stockroom. I assembled and tested the atomizers using the work card and graphics manual as a reference. As part of the testing of atomizers, they are proof leak tested and flow tested from an input pressure of 60 psi up to 450 psi. I did not see any leaks as part of the test. Not all passed test due to incorrect flow volume or streaking patterns in the spray angle check. Approximately 7 passed and I made them svcable per air carrier wyy (unsvcable tag) and then on to air carrier abc (svcable form). The assembled parts are kept in the immediate area, in a parts cabinet to be used in the assembly of the burner can assembly. The following week (jul/mon-fri/00) I was on vacation. On jul/mon/00, my first day back from vacation, a fellow employee came up and showed me an atomizer assembly that I had assembled on jul/fri/00 that the flow divider part #3883998-1 had cracked off from where it threads into the atomizer housing. Upon closer inspection with fellow employees experienced in the assembly of atomizers, we were never able to come up with a reason except that I may have inadvertently overtorqued the flow divider that threads into atomizer housing. Later in the week, 3 each additional flow dividers were found to be cracked off. By this time we got air carrier engineers involved with not a clear cause as to what was causing the cracking. The tooling fixture that holds the housing while assembling the atomizers in 3 studs in an aluminum block that is clamped in a vise. There is not a means of securing the housing to the fixture, it just rests on the 3 studs. With this in mind, I don't feel I could have overtorqued the flow dividers to the extent of causing them to crack off because the assembly would have come up out of the fixture. This information was shown in detail to air carrier engineers to the extent that he witnessed the removal of the flow divider from 1 of the 11 atomizer assemblies in question. By now both my supervisor and engineering were aware the problem existed and they knew it could cause additional problems. I tracked down the location of the atomizers in question with 2 each still in air carrier stores and some in outlying cities that air carrier does burner can changes. I received back the 2 each from air carrier stock room in ZZZ. I disassembled 1 atomizer assembly and found that it had no crack indications using a 10X-power magnifying glass. I disassembled the other and it had a crack using the 10X magnifying glass. By this time the supervisor and engineering people went home for the day. The supervisor called in later on and I spoke with him that I had found another atomizer assembly that was in a burner can assembly from stock that was cracked. The engineer told me the flow dividers were to be manufactured of a steel alloy. Air carrier engineers were in the process of sending to a lab to find out the composition and the heat treating qualities of the cracked flow dividers. It appears the quality of metal in the manufacturing of the flow dividers is also in question. Aside from what the lab reports find, it is of the utmost importance to follow assembly and torque requirements as they are spelled out in the work card. Callback conversation with reporter revealed the following information: the reporter stated the flow dividers were all recovered and the manufacturer has made immediate changes to the torque of the flow divider into the body of the atomizer. The reporter said the torque was lowered from 120 inch pounds to 40 inch pounds. The reporter said the air carrier APU engineering staff discovered the flow divider metal was much too hard and will be corrected by the manufacturer. The reporter stated that boeing and the manufacturer are all alerted to the problem and the rest of the industry using this model of APU will be advised.

Google
 

Original NASA ASRS Text

Title: IT WAS DISCOVERED THAT APU'S USED ON B737-200 -300, AND -400 SERIES ACFT HAD BURNER CAN ATOMIZER FLOW DIVIDERS CRACKING.

Narrative: I WORK FOR ACR AS A MECH IN THEIR APU SHOP IN ZZZ, US. ON JUL/FRI/00 I WAS ON A POST SHIFT OF 4 HRS OF OVERTIME. I WAS ASSIGNED TO ASSEMBLE AND TEST ATOMIZER ASSEMBLIES. THEY ARE ALLIED SIGNAL ASSEMBLY PART #36-1193-8, THAT ARE USED ON BURNER CAN ASSEMBLIES THAT ARE INSTALLED ON ALLIED SIGNAL APU'S GTCP85-129C, -129E, -129H MODELS. ACR HAS A WORK CARD THAT COVERS THE ASSEMBLY AND TEST OF THE ATOMIZER ASSEMBLY, WORK CARD #WXXX. I HAVE ASSEMBLED ATOMIZER APPROX 3 OTHER TIMES, BUT THIS WAS THE FIRST TIME I ASSEMBLED THIS MODEL. THE HOUSINGS, APPROX 11 EACH, WERE NOW OUT OF STOCK. THE PART IN QUESTION, THE FLOW DIVIDERS (ALLIED SIGNAL PART #3883998-1) WERE ALSO NEW FROM THE COMPANY STOCKROOM. I ASSEMBLED AND TESTED THE ATOMIZERS USING THE WORK CARD AND GRAPHICS MANUAL AS A REF. AS PART OF THE TESTING OF ATOMIZERS, THEY ARE PROOF LEAK TESTED AND FLOW TESTED FROM AN INPUT PRESSURE OF 60 PSI UP TO 450 PSI. I DID NOT SEE ANY LEAKS AS PART OF THE TEST. NOT ALL PASSED TEST DUE TO INCORRECT FLOW VOLUME OR STREAKING PATTERNS IN THE SPRAY ANGLE CHK. APPROX 7 PASSED AND I MADE THEM SVCABLE PER ACR WYY (UNSVCABLE TAG) AND THEN ON TO ACR ABC (SVCABLE FORM). THE ASSEMBLED PARTS ARE KEPT IN THE IMMEDIATE AREA, IN A PARTS CABINET TO BE USED IN THE ASSEMBLY OF THE BURNER CAN ASSEMBLY. THE FOLLOWING WEEK (JUL/MON-FRI/00) I WAS ON VACATION. ON JUL/MON/00, MY FIRST DAY BACK FROM VACATION, A FELLOW EMPLOYEE CAME UP AND SHOWED ME AN ATOMIZER ASSEMBLY THAT I HAD ASSEMBLED ON JUL/FRI/00 THAT THE FLOW DIVIDER PART #3883998-1 HAD CRACKED OFF FROM WHERE IT THREADS INTO THE ATOMIZER HOUSING. UPON CLOSER INSPECTION WITH FELLOW EMPLOYEES EXPERIENCED IN THE ASSEMBLY OF ATOMIZERS, WE WERE NEVER ABLE TO COME UP WITH A REASON EXCEPT THAT I MAY HAVE INADVERTENTLY OVERTORQUED THE FLOW DIVIDER THAT THREADS INTO ATOMIZER HOUSING. LATER IN THE WK, 3 EACH ADDITIONAL FLOW DIVIDERS WERE FOUND TO BE CRACKED OFF. BY THIS TIME WE GOT ACR ENGINEERS INVOLVED WITH NOT A CLR CAUSE AS TO WHAT WAS CAUSING THE CRACKING. THE TOOLING FIXTURE THAT HOLDS THE HOUSING WHILE ASSEMBLING THE ATOMIZERS IN 3 STUDS IN AN ALUMINUM BLOCK THAT IS CLAMPED IN A VISE. THERE IS NOT A MEANS OF SECURING THE HOUSING TO THE FIXTURE, IT JUST RESTS ON THE 3 STUDS. WITH THIS IN MIND, I DON'T FEEL I COULD HAVE OVERTORQUED THE FLOW DIVIDERS TO THE EXTENT OF CAUSING THEM TO CRACK OFF BECAUSE THE ASSEMBLY WOULD HAVE COME UP OUT OF THE FIXTURE. THIS INFO WAS SHOWN IN DETAIL TO ACR ENGINEERS TO THE EXTENT THAT HE WITNESSED THE REMOVAL OF THE FLOW DIVIDER FROM 1 OF THE 11 ATOMIZER ASSEMBLIES IN QUESTION. BY NOW BOTH MY SUPVR AND ENGINEERING WERE AWARE THE PROB EXISTED AND THEY KNEW IT COULD CAUSE ADDITIONAL PROBS. I TRACKED DOWN THE LOCATION OF THE ATOMIZERS IN QUESTION WITH 2 EACH STILL IN ACR STORES AND SOME IN OUTLYING CITIES THAT ACR DOES BURNER CAN CHANGES. I RECEIVED BACK THE 2 EACH FROM ACR STOCK ROOM IN ZZZ. I DISASSEMBLED 1 ATOMIZER ASSEMBLY AND FOUND THAT IT HAD NO CRACK INDICATIONS USING A 10X-PWR MAGNIFYING GLASS. I DISASSEMBLED THE OTHER AND IT HAD A CRACK USING THE 10X MAGNIFYING GLASS. BY THIS TIME THE SUPVR AND ENGINEERING PEOPLE WENT HOME FOR THE DAY. THE SUPVR CALLED IN LATER ON AND I SPOKE WITH HIM THAT I HAD FOUND ANOTHER ATOMIZER ASSEMBLY THAT WAS IN A BURNER CAN ASSEMBLY FROM STOCK THAT WAS CRACKED. THE ENGINEER TOLD ME THE FLOW DIVIDERS WERE TO BE MANUFACTURED OF A STEEL ALLOY. ACR ENGINEERS WERE IN THE PROCESS OF SENDING TO A LAB TO FIND OUT THE COMPOSITION AND THE HEAT TREATING QUALITIES OF THE CRACKED FLOW DIVIDERS. IT APPEARS THE QUALITY OF METAL IN THE MANUFACTURING OF THE FLOW DIVIDERS IS ALSO IN QUESTION. ASIDE FROM WHAT THE LAB RPTS FIND, IT IS OF THE UTMOST IMPORTANCE TO FOLLOW ASSEMBLY AND TORQUE REQUIREMENTS AS THEY ARE SPELLED OUT IN THE WORK CARD. CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: THE RPTR STATED THE FLOW DIVIDERS WERE ALL RECOVERED AND THE MANUFACTURER HAS MADE IMMEDIATE CHANGES TO THE TORQUE OF THE FLOW DIVIDER INTO THE BODY OF THE ATOMIZER. THE RPTR SAID THE TORQUE WAS LOWERED FROM 120 INCH LBS TO 40 INCH LBS. THE RPTR SAID THE ACR APU ENGINEERING STAFF DISCOVERED THE FLOW DIVIDER METAL WAS MUCH TOO HARD AND WILL BE CORRECTED BY THE MANUFACTURER. THE RPTR STATED THAT BOEING AND THE MANUFACTURER ARE ALL ALERTED TO THE PROB AND THE REST OF THE INDUSTRY USING THIS MODEL OF APU WILL BE ADVISED.

Data retrieved from NASA's ASRS site as of July 2007 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.