Time Saving in the X-Ray Production Control by using ASD

Transcription

1 ECNDT 2 - Poster Time Saving in the X-Ray Production Control by using ASD Elena TOSTI, AVIO Propulsione AeroSpaziale, Colleferro, Italy Abstract. Time reduction in production controls are very important to avoid long waiting time between phases. This either in large number production of middle dimension units and in small number production of large dimension units. Both cases exist in the AVIO Space Division in Colleferro. In order to reduce X-Ray time control, the use of ASD has been implemented with a time saving better of 7% now guaranteed. Introduction In a production cycle the product undergoes Non Destructive Tests. The NDT are important for monitoring of the quality of the product after a production phase and eventually to submit the article to repairs or to reject in case of irreparable faults that compromise the functioning of the object. These reasons justify the costs joined with NDT. Nevertheless NDT sites are often heap bay for the incoming units since control time is not lined up with time of the previous phase and it is in general expensive duplicate the control process. For an organisation of the production line following the concept of lean manufacturing the requirements are to smooth technical times at each bay below the takt time without losses in effectiveness. In view of cost reduction of the product, improvements in NDT which imply profits, even with investments, are taken into great consideration. This paper presents the solution taken in AVIO to reduce costs of NDT mainly by shortening the control time. 1. NDT in Production Activities 1.1 Organisation of the Production Lines The production line organisation following the concept of lean manufacturing (see Fig. 1) considers two main requirements, namely: technical times smoothed below the takt time (in Fig. 1 the arrow indicates a phase out the takt time characteristic of the production line in consideration) and manpower organised through working team. These two requirements strongly also affects the logistical aspect of the production line. The first step minimising the crossing time of the item is to conceive a plant dedicated to the production of this item where materials come in and the final piece goes out. In the production plant dedicated to the item all manufacturing phases takes place: realization of the roughing piece, mechanical finish, resistance tests, NDT controls, dimensional tests, final assembling. In the dedicated plant the item moves in agreement with the manufacturing phase flow. In each processing bay the item stands just the time to perform the process. The technical time of each processing phase does avoid a pile up of pieces in the bay. In the 1

2 case a pile up is generated (see arrow in Fig. 1), means that the technology relevant to the process is not lined up with the previous one in the production cycle. When wait-time characterizes a production phase it is necessary to critically analyse the process and to initiate corrective actions to remove the bottleneck effect in the production flux. THERMAL PROTECTION APPLICATION THERMAL PROTECTION POLIMERISATION IN AUTOCLAVE METALLIC PARTS PREPARATI ON MECHANICAL FINITURE MANDREL ASSEMBLING & DI- SASSEMBLING FILAMENT WINDING METALLIC PARTS ASSEMBLING TIME [hours] TAKT TIME IN HIDRO-PROOF TEST X-RAY DIMENSIONAL TEST OUT PHASE Figure 1 Lean manufacturing production line and takt time 1.2 Smoothing of NDT Time Cut down production costs means to upgrade technologies in view of shortest time and highest reliability. NDT also are involved in these improvements. To this purpose, after the selection of NDT methods that meet the reliability requirement, techniques allowing shortest control time shall be designed. In particular highly manual old techniques shall be avoided because too slow. In pursuit of optimal NDT for production rate, several methods have been tested in AVIO among conventional and not conventional NDT: x-ray, US, IRlock-in, shearography, acoustic emission. For the time being, the results of these tests, the discussion of which are beyond the subject of this paper, show that concerning the product type in AVIO, the radiographic control has the highest reliability. For this reason activities focused in the reduction of x-ray control time have been initiated in AVIO by means of the Six Sigma methodology. The results of these activities are summarised hereinafter. 2. Product Type and Relevant NDT in AVIO AVIO Space Division in Colleferro is involved in the construction of solid propellant rockets used on European launchers (Ariane 3,, and VEGA). AVIO Space Division is also involved in the construction of solid propellant rockets of tactical missiles in defence cooperations among European governments. The last generation of solid propellant rockets designed in AVIO uses case in carbon fibre instead of metal. In both applications: space and tactical propulsion, the case construction technology is the same, 2

3 but the dimension of the final product much different (see Fig. 2). 388 mm mm 932 Figure 2 Typical dimensions for solid propellant thrusters used in space application (on the left) and in tactical application (on the right) In this type of rockets: the possible defects, the roughness of the surface, the shape, the existence of not accessible areas, the internal feature of the case obtained by means of complex layering operations in filament winding technology, make NDT control not easily feasible. The only NDT method allowing for unambiguous results is the radiographic control. Unfortunately the severe requirement in the defect detectability (thin delaminations how.3 mm inside mm of material thickness) needs to use conventional film and exclude videoscopy systems already available in AVIO since 198 for Ariane rockets x-ray control. Radiographic controls using conventional film fulfil the sensitivity requirement, but not the shortest control time requirement that causes the bottleneck effect in the production flux either in large number production of middle dimension units and in small number production of large dimension units. The RT using conventional film implies several causes of lengthening of control time: exposure preparation in the dark room and in the exposure area, exposure time, film processing. The Fig. 3 shows the typical (theoretical) time for each operation. In case of units of large dimensions, being implied high energy accelerator (9-1 MeV), the phase of film positioning/removing can require time longer than in film processing. The long time is due to safety operations implied when high energy accelerator is used. The processing time indicated in Fig. 3 is at least 1 minute (time necessary to the film insertion in the processing machine), but 8 minutes have to be considered if is necessary the control of the result before the next radiography. All causes of lengthening of the control time give a serious percentage increase of the overall time that only the imaging technique in x-ray can overcome. 3

4 CYCLE DARK ROOM BUNKER ROOM CHASSY LOADING POSITIONING 1 EXPOSURE Te Te DENSITY CHECK.1.1 SHELVING PROCESSING REMOVING 1 BUNKER ACCESS ABILITATION 1-3 READING: DEFECT DIMENSIONING APCEPTANCE LEGENDA PHASE NOT INCLUDED IN TIME REDUCTION DESIGN MANUAL AND AUTOMATIC OPERATIONS REQUIRED TIME [min] FOR RX TUBES AND ACCELERATORS Figure 3 Time required for each operation of an x-ray control through conventional film. Shortening of Control Time by Using Digital Flat Panel Radiographic Systems Among the imaging systems, the x-ray Amorphous Silicon Detector (ASD) has been selected as its performance reach the radiographic class C3 (EN 8.1) which is the class used for control in AVIO and it allows for control without access to the inspection area. Performance was tested both on middle dimension units and large dimension units. The results [1] obtained through x-ray ASD, showed the same or better image quality with respect conventional film. The first digital radiographic system implemented in AVIO Colleferro was for control of middle dimension units and it is composed by (see Fig. ): X-ray tube 22 kv, W (1 mm focal spot) and control unit X-ray tube diaphragm and control unit DR flat panel 127 μm pixel pitch, Gd 2 O 2 S conversion screen, 29.3x. cm pixel area Image processing and remotely control unit Motorised movements test bench and remotely control unit Surveillance camera Control by means x-ray ASD is performed in agreement to the flow in Fig.. In the figure are also specified typical time (theoretical) for each operation. The comparison with Fig. 3 lets to determine the advantages in time saving when x-ray ASD is used. Time saving is consequent to remotely performed operations avoiding continuous

5 access to the inspection area, as well as to shorter exposure times (see Table 1). Exposure parameters by using ASD detector are smoother with respect conventional film not only concerning exposure time but also energy. This since ASD detector efficiency is higher than the film efficiency. ROOM BUNKER SURVEILLANCE CAMERA X-RAY ASD INTERNAL VIEWS X-RAY TUBE DIAPHRAGM UNIT X-RAY TUBE TEST BENCH UNIT X-RAY TUBE UNIT Figure X-ray control system using ASD detector ASD CYCLE ROOM FLAT PANEL POSITIONING IMAGE PROCESSING. IMAGE STORAGE EXPOSURE IN LIVE ACQ RX TUBE SETTING.3.3 IMAGE ANALISYS: COLLIMATOR SETTING defect dimensioning acceptance 1 FRAMES INTEGRATION.. LEGENDA PHASE NOT INCLUDED IN TIME REDUCTION DESIGN REQUIRED TIME [min] FOR RX TUBES OPERATIONS FROM REMOTE Figure Time required for each operation of an x-ray control through ASD

6 ZONE of THE MOTOR TABLE I Exposure parameters used for X-ray control of middle dimension solid propellant thrusters by means conventional film and x-ray ASD ENERGY [kv] INTENSITY [ma] EXPOSURE TIME [s] No. of EXPOSURES ( ) [#] OTHER PARAMETERS ASD ASD ASD ASD ASD DOME ASD:Fine Film: FO = CYLINDER x x Class C3 1 mm 3 mm FLANGE AT CYLINDER END (double exposure) FO = 12 mm 19 mm Radiographic density: ASD:Fine FO = 1 mm 3 mm FO = mm 3 mm METALLIC CONNECT- ION THE CYLINDER 18 (*) (*) 3 (*) 3x / (*) exposures of sectors on the cylinder let the simultaneous control on both: fibre case and metallic connections, due to ASD higher dynamic than film [2]. In order to evaluate the true time reduction in replacing the x-ray control by conventional film with x-ray control using ASD, two lots of 2 thrusters each have been considered and the relevant distributions of the control times for unit have been traced out (see Figs. A and B). In Fig. A the control times necessary when traditional films are used, are in any case longer than the takt time. On the contrary, the control times are greatly reduced below the takt time when the control is performed through x-ray ASD. The times needed for control by means ASD have been lowered below the required specifications, but the technique allows control time below the monitored mean value of 1.8 hours. In effect the technical time for frame is 2 minutes (see Fig. 7) which is the minimum got time. In this time is excluded the analysis of the frame necessary for dimensioning of eventual defects and for acceptance because this operation can be performed by a second operator on a parallel PC. The same procedure apply to conventional film, where other operators evaluate defects during x-ray control. In Fig. 7, where the distribution of the time for single frame is shown, the empty rectangles are relevant the time typical for a skilled operator, while the full rectangles are relevant the time typical for not skilled operator or obtained when the analysis is performed during the control. In considering the 2 minutes or technical time for frame, the evaluated technical time for control of the thruster is 1 hour with x-ray ASD versus.3 hours when conventional films are used for control. This improvement, not actually necessary, can be considered for future demands.

7 8 FREQUENCY TAKT TIME TIME [hours] Observed Performance Exp. "Within" Performance Exp. "Overall" Perfor PPM< LSL PPM < LSL 9 PPM< LSL A 1. Process Capability Analysis for TEMPI RX (DR TIME [hours] Figure Distributions of x-ray control time per unit by means film (A) and ASD (B) FREQUENCY Observed Performance Exp. "Within" Performance B PPM < LSL PPM < LSL PPM < LSL TAKT TIME. Exp. "Overall" Performa 18 1 FREQUENCY TIME FOR SINGLE IMAGE [hours] Figure 7 Distributions of times for single image by x-ray ASD. Conclusions The use of x-ray control by means ASD in place of conventional film, has allowed the reduction of the control time in the production line for middle dimensions thrusters. The reduction has been 7% and a potential reduction up to 8% has been evaluated how the maximum allowed. Acknowledgements The author wishes to acknowledge Dr. U. Heike, Mr. D. De Toni and Mr. M. Calvi for their help during the set-up of the system. References [1] Tosti E., Gatta A., Controlli non Distruttivi effettuati su Componenti Aerospaziali mediante un Sistema di Radiografia Digitale, 1 Congresso Nazionale AIPND, Ravenna 23, ITALY. [2] Ewert U., Zscherpel U., and K. Bavendiek Film Replacement by Digital X-Ray Detectors - The Correct Procedure and Equipment, 1 th WCNDT, Montreal 2, CANADA. 7