PSI - Issue 60

Sreerag M N et al. / Procedia Structural Integrity 60 (2024) 20–35 Sreerag M N/ Structural Integrity Procedia 00 (2023) 000 – 000

34

Since repair is expected to cause local distortions in the geometry, these distortions (also called as profile error) were quantified for estimating the stress field in this repair zone. Axisymmetric FE model with measured profile error data is used for the failure pressure assessment. (Similar to details given in section 4.2). Unflawed failure pressure was estimated at all the locations, where the profile error data was available. Both the models were taken i.e. full model with R2 weld properties as well as the model having parent metal properties at regions away from weld with R2 weld properties near the weld. Using this unflawed burst pressure, failure pressure in presence of flaw (5x2 mm) is estimated. Details are summarized in Table 12 given below.

Table 12: Comparison of PPT measured strain and prediction at R2 weld location Cylindrical shell outer surface

Cylindrical shell Inner surface

Ch. No.

Predicted strain (µ ɛ ) Measured strain (µ ɛ )

Ch. No.

Predicted strain (µ ɛ )

Measured strain (µ ɛ )

39-L 40-H 41-L 42-H 43-L 44-H 51-L 52-H 53-L 54-H 55-L 56-H

1457 6101 1194 6090 1018 6101 1195 6082 1311 6081 1195 6082

1770 6279 1373 6335 1224 6227 1269 6128 1614 6009 1354 6082

45-L 46-H 47-L 48-H 49-L 50-H 57-L 58-H 59-L 60-H 61-L 62-H

1283 6146 1170 6137 1910 6125 1631 6127 1175 6127 1631 6127

1155 5792 1140 5611 1730 5793 1843 5773 1187 5849 1536 5849

8. Conclusion

 Numerical study of the effect of weld repair and repair length on the burst pressure of solid rocket motor case was studied. From this study it is found that, structural assessment with weld repair mechanical properties for all the domains of the FE model (i.e. parent metal region as well as weld region) yields a conservative estimate of failure pressure. This is true for long seam weld as well as cirseam weld.  In case of long seam welds, the difference in estimated failure pressure with and without using the weld properties is of 1%. But in case of cirseam welds this difference is as high as 20%. This usually results in low design margins (and sometimes negative margins also) over ultimate design burst pressure. In such cases it is better to estimate the failure pressure considering both parent metal as well as the weld properties.  The effect of weld repair length in cirseam weld and long seam weld on the failure pressure is separately studied. In long seam welds, as the repair length (i.e. in axial direction) increases, reduction in the failure pressure is observed. However, in cirseam welds, as circumferential weld repair length increases, no appreciable reduction in failure pressure is observed.  The actual hardware which had weld repair has successfully undergone proof pressure test. During proof pressure test the hardware was strain gauged at the repair location. The predicted strains were closely matching with the measured strains. This hardware was successfully flown in one of the LVM3 mission. This invariably proves the methodology described this paper. However, no specimen testing was done for evaluating the effect of weld repair length.  A typical case study for salvaging the second weld repair deviation in the cirseam weld is presented here. Acknowledgements We express our sincere thanks and gratitude towards EFA/MME and QRFG/SR teams for sharing the weld repair details in the case study presented in this paper.