Failure analysis of automotive battery parts

Failure analysis of automotive battery parts

Engineering Failure Analysis 16 (2009) 2217–2223 Contents lists available at ScienceDirect Engineering Failure Analysis journal homepage: www.elsevi...

645KB Sizes 0 Downloads 18 Views

Engineering Failure Analysis 16 (2009) 2217–2223

Contents lists available at ScienceDirect

Engineering Failure Analysis journal homepage:

Failure analysis of automotive battery parts J. López, R. Navarro, J.M. Gallego, F. Parres, S. Ferrándiz * Polytechnic University of Valencia, Department of Materials and Mechanical Engineering, 03801 Alcoy, Spain

a r t i c l e

i n f o

Article history: Received 3 February 2009 Accepted 1 March 2009 Available online 9 March 2009 Keywords: Plastics failure Failure in processing Stress analysis

a b s t r a c t In the injection molding process defective pieces are sometimes detected, showing weak points which break easily. In this work, the possible causes of these failure points are evaluated. Thermal analysis techniques were applied to the material to better understand the degradation that takes place with relation to the transformation process. Simulations were made to determine the possible causes of deterioration and at the same time reproduce the effects of this fragility. The creation of the weak points was simulated and their origins were determined. Ó 2009 Elsevier Ltd. All rights reserved.

1. Problem definition All through the manufacture of polypropylene battery bodies for the car industry, many pieces are detected with extreme fragility during their everyday use or even when dropped during handling. The defects in these battery bodies caused their production rejection, as they did not meet the minimum quality requirements. Moreover, the detection of these faulty products caused, on many occasions, production of the battery bodies to be stopped. Taking as a starting point the conditioning factors already mentioned, we propose three possible causes of the fragility of these injection pieces: design, process and/or material. The design of the piece in question can be discarded immediately, as the piece has been manufactured for several months without any fragility problems. The material can also be described as a cause, as it is being used in other products without problems. Therefore the only possibility that remains is that the weak points in the battery bodies come as a consequence of the transformation process. Initial visual analysis of the defective pieces allowed us to see that the break always occurred in the same area, which was close to the one of the injection points. The rest of the body showed no indication of fragility or weak points in the material. With the aim of solving this problem, various samples taken from different areas of the battery body were analysed using differential scanning calorimetry. The thermal analyses carried out showed differences in thermic stability between the areas which were sound and the defective areas. This result meant that an exhaustive analysis of the mold used for polypropylene was necessary. The results of this analysis showed that one of the injection channels was partially blocked. This blockage caused an increase in the shear stress on the injected material, which in turn led to greater degradation of the material delivered through it. This hypothesis was reproduced and confirmed by means of the use of the injection process simulation software. 2. Background The automotive industry is subjected to high specifications with the companies who supply the different parts that make up the vehicle. Given the level of rivalry in this sector, the auxiliary industries around this sector must produce parts with * Corresponding author. Tel./fax: +34 966528422. E-mail address: [email protected] (S. Ferrándiz). 1350-6307/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.engfailanal.2009.03.004


J. López et al. / Engineering Failure Analysis 16 (2009) 2217–2223

Fig. 1. Automotive battery analysed.

three basic characteristics: high productivity, low price and high quality [1]. If companies in these auxiliary industries do not fulfill these requirements, then they will lose clients (car manufacturers). When a problem arises in any piece destined for car manufacture, it must be identified and corrected extremely quickly. In this study, the polymer body of a car battery is analysed (Fig. 1). This piece, produced in a polypropylene polymer, is an injection molded piece with six injection points. An initial visual analysis identified a zone of high fragility close to one of the injection points, which caused the appearance of different length cracks in the area mentioned (Figs. 2 and 3). This phenomenon was detected in an injection line and randomly affected part of the production. The methodology employed to study this defect in the battery consisted of accomplish diverse analyses of the design, material and manufacture process [2] (see Fig. 4). Initially, the design of the different morphologies presented by the battery was discounted as a factor in their increased fragility, as the defect only appeared in one of the production lines and therefore the design can be considered to be correct. The second study to be carried out was on the characterisation of the material being used. This too could be immediately discounted for the same reasons as design faults. In despite of this, it seemed interesting to execute some tests using thermic analysis techniques, in which various areas of the battery, both fragile and sound, were compared. 3. Results and discussion Calorimetric analysis was performed using DSC Mettler-Toledo 821 equipment (Mettler-Toledo, Schwerzenbach, Switzerland). Samples of weight between 6 and 7 mg were used. To determine variations in the level of crystallinity of the samples a first heating (30–250 °C at 10 deg min l 1) was completed, followed by a cooling process (250–30 °C at 10 deg min l 1) to eliminate the thermal history, and was finished with a second heating (30 °C at 200 °C at 10 deg min l 1). The tests were performed in a nitrogen environment (flow rate 50 ml min 1). Finally, to determine the degree of degradation taking place, the conditions of the thermic analysis were varied, in this case using a single thermic ramp which began at 30 °C and finished at 250 °C at a rate of 10 °C min 1, but in this case the

Fig. 2. Area of breakage.

J. López et al. / Engineering Failure Analysis 16 (2009) 2217–2223


Fig. 3. Close-up of the area of breakage.

Fig. 4. Process analysis tree.

inert atmosphere caused by nitrogen was removed in order to study the sensitivity of the polypropylene to oxidation in the different areas studied. The crystallinity analysis results obtained from the fusion energies did not show any differences between material analysed from broken areas and accurate areas [3,4]. The results suggest that the material used is in good condition. In both cases the first peak (Fig. 5), corresponding to the absolute value of the fusion energy, is less than the crystallisation energy obtained in the endothermic peak from the cooling process. It is also less than that obtained from the second heating cycle, after a slow cooling. Therefore the mold cooling system acted correctly in both areas, and we can discount differential freezing as a cause of the fragility. In contrast, when we analysed the material’s degradation behaviour, there was a clear difference between the materials. The material from the fragile zone oxidised much more quickly than the material from areas without problems [6]. Material from the fragile zone oxidised at 219 °C compared with 230 °C for the sound material (Fig. 6). This was certainly the cause of the fragility and so it was necessary to establish on what was causing degradation in the material during the injection process. Finally, it only remained to analyse the transformation process as the cause of the problem. If the origin of the problem is a variation in the cooling of the battery, it could be due to a problem in the cooling circuits/cycles, while a degradation problem would mean a revision of the equipment. A simulation of the injection of the piece was used to identify which hypothesis was correct. Three simulations were initially programmed (Table 1). The first simulation was carried out under normal


J. López et al. / Engineering Failure Analysis 16 (2009) 2217–2223

Fig. 5. DSC curves of different zones of automotive battery.

Fig. 6. Thermal degradation.

transformation conditions. The second was performed with a change in the refridgeration conditions, in which we created thermic differentials which would in turn cause residual tensions. The third simulation was achieved with changes in the external conditions of the process. For this, physical changes in the piece were incorporated to emulate restrictions in the flow of polymer and thus an increased shear stress tension [5,7,8]. The first simulation, Fig. 7, was accomplished under normal process conditions. We can see shear stress values in the injection points which in no case pass the values permitted for the material, which in this case is 29 MPa. This means that in a normal transformation, there should not be breaks caused by residual stress [9]. In the second simulation, the effects of an unbalanced cooling were analysed. A cooling circuit was simulated which acted on half a piece. Temperatures over 10 °C were reached [10]. The analysis of the stress generated was correlated with the fault observed. Quantitatively, the tensions were generated in a uniform way for all injections of material, Fig. 8, and qualitatively the maximum stress values achieved were 29 MPa. If we consider that the study was carried out on a hypothetical fault in the refridgeration system, this can be discarded, given that the data generated by the simulation does not correspond with the fault observed.

J. López et al. / Engineering Failure Analysis 16 (2009) 2217–2223

Fig. 7. Simulation under normal conditions.

Fig. 8. Thermic jump created.



J. López et al. / Engineering Failure Analysis 16 (2009) 2217–2223

Fig. 9. Shear stress values with reduction of gates.

Fig. 10. Residual stress evolution throughout the thickness of the piece.


J. López et al. / Engineering Failure Analysis 16 (2009) 2217–2223 Table 1 Shows the values used to simulate the injection. Material data: polymer

Moplen EP548T: Basell polyolefins Eu

PVT model Specific heat (Cp) Thermal conductivity Transition temperature

2-Domain modified Tait 2960.0000 J/kg C 0.1600 W/m C 130.0000 °C

Machine parameters Maximum machine clamp force Maximum injection pressure Maximum machine injection rate Machine hydraulic response time

7.0002E+03 tonne 1.8000E+02 MPa 5.0000E+03 cm3 s 1.0000E 02 s

Process parameters Fill time Injection time has been determined by automatic calculation Stroke volume determination Cycle time Velocity/pressure switch-over by Packing/holding time


1.1000 s Automatic 35.0000 s Automatic 10.0000 s

The third simulation model showed us a break that was caused because the shear stress values permitted were surpassed (29 MPa). The way these values were reached was in a drastic reduction in the flow through two of the gates in this area, and then to observe the effects. The simulation in Fig. 9 showed behaviour very similar to that observed in the real piece. The values throughout the inlet surface of the material of the piece have reasonable values and are always below the maximum tension permitted for the material. Quantitatively and qualitatively the results agree. In the break zone, there is a tension profile which surpasses 29 MPa, not only on the exterior of the piece (the face which touches the mold), but also in the interior. In a case where there was no narrowing, we would have a tension profile in the piece with a maximum of 8 MPa, much lower than the established limit. The maximum value permitted for the material is 29 MPa, and as we can see in Fig. 10, 35 MPa was reached. This must be interpreted as a brittle break at any moment. 4. Conclusions In the search for a quick solution to a problem of production, the characterisation of the polymer using efficient differential calorimetry techniques provides information on the origin of the fault. In this case, crystallinity analysis did not show differences between the fracture zone and the rest of the piece. As a second stage, using the data we obtained we can use the simulations as a tool which allows us to confirm the experimental results and offer a solution. In this case, faults which affect cooling and material degradation were simulated. Failures due to cooling system do not correspond to the type of fracture observed, while the data obtained from the section restrictions do correspond to this crack in the real problem. This data led us to concentrate on a revision of the mold in the areas previously mentioned and thus rectify the situation in a short time. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

Castro Paulo MST, Fernández AA. Methodologies for failure analysis: a critical survey. Mater Des 2004;25:117–23. Ezrin M, Lavigne G. Unexpected and unusual failures of polymeric materials. Eng Fail Anal 2007;14:1153–65. Wintle JB, Pargeter RJ. Technical failure investigation of welded structures. Eng Fail Anal 2005;12:1027–37. López J, Balart R, Sánchez L, Nadal A. Application of thermal analysis to the study of recycled polymer materials in blends and degradation. Polym Anal Degr; 2000 [ISBN 1-56072-861-2]. Shoemaker Jay. Moldflow design guide: a resource for plastics engineers. Hanser, Cop; 2006 [ISBN 1569904030]. Loyd E, White JR. Photo-sensitivity of recycled photo-degraded polypropylene. J Mater Sci Lett 2002;21:627–9. Dunn Russell F, McSwain Richard H, Terry Mills, Brain Malone. Failure of plastic press release buttons in automobile seat belts. Eng Fail Anal 2005;12:81–98. Moura GCR, Aguilar MTP, Pertence AEM, Cetlin PR. The failure analysis of a deep drawing dye in the manufacturing of an automotive shock absorber cap. Eng Fail Anal 2004;11:943–50. Krishnamachari Sadagopa I. Applied stress analysis of plastics. A mechanical engineering approach. Springer; 1994. Wright David. Failure of plastics and rubber products: causes, effects and case studies involving degradation. Rapra Technology; 2001 [ISBN 185957517X].