·Home ·Table of Contents ·Aeronautics and Aerospace

Characterisation of defects in Large Solid propellant Rocket Motors K.Venkata Rao, B.Munirathinam, C.Subbiah, P.V.Sai Suryanarayana, & K.Viswanathan SPROB / SHAR Centre, Sriharikota, India - 524 124. Contact

1.0 Introduction

In the modern era of (internet) communications, there are consistent demands from various countries of the globe for launching higher pay-load satellites into either lower earth orbits (LEOs) or geo-synchronous orbits (GSOs). This puts considerable pressure on the manufacturers of rocket motors to design and manufacture increasingly higher capacity and efficient solid propellant boosters to generate the necessary thrust at the initial stages of flight. The production of defect-free large solid propellant rocket boosters however is a formidable challenge and it calls for stringent quality control measures at various stages of its processing. Due to the inherent complexities of large solid rocket motor (SRM) manufacture, defects may crop into the final product, despite strict vigil on the quality control during processing.

Due to inherent limitations and complementary nature of different types of NDT methods, many of them need to be applied depending on their suitability to each rocket motor at various stages of its production. In the inspection of a solid propellant rocket motor for instance,

• NDT methods such as radiography, ultrasonic testing and dye-penetrant testing are being employed for the inspection of rocket case hardware,
• longitudinal normal beam ultrasonic pulse-echo method is applied for checking the bond-line integrity of case and insulation layer and
• X-ray radiography is employed for evaluating both the integrity of propellant mass and the bond-line integrity of propellant and insulation.
For the inspection of smaller size rocket motors, conventional X-ray generators up to 450 KV are sufficient to meet the required 1-1T quality level. The radiographic evaluation of larger size rocket motors with diameters ranging from 1000 to 2800 mm however requires the use of high energy linear accelerators (linac) with electron energies ranging form 9 to 15 MeV to attain the same or better quality level. Apart from the above, specialized radiographic techniques have to be developed to properly quantify the propellant defects and qualify a processed rocket motor.

2.0 SRMs and Radiography methodology

A typical SRM is made by vacuum casting a homogeneously mixed propellant slurry into an insulated case and curing at an elevated temperature. Due to various reasons possibility of ending up with a SRM containing defects can not be ruled out. Therefore, a thorough NDT inspection becomes indispensable to assess the suitability of a SRM for a specific mission. Typical defects that may occur are

• propellant defects such as cracks, tears, voids, porosity, foreign material inclusion and port deformation
• interfacial defects such as debonds, delaminations, and resin filling defects.
The detection of these defects is very essential as these may lead to unacceptable pressure levels during the SRM performance and might lead to failure of the mission itself.

Two types of radiographic techniques have been standardized to detect and locate the above mentioned defects in SRMs by interfacial coverage, and propellant coverage. The interfacial coverage is for revealing debonds if any between interfaces such as case / insulation / propellant and defects if any in the peripheral part of the propellant close to insulation. To achieve this, X-rays are directed tangentially to the above interfaces and the emerging radiation is recorded on a film as shown in fig.1 (a). The SRMs are to bekept horizontally or vertically for this inspection. The propellant coverage is provided to assess the quality of propellant mass. In this coverage, X-rays are directed normal to the propellant geometry, as in fig. 1 (b).

Fig 1(a): Interfacial Radiography Set-up

Fig 1(b): Propellant Radiography Set-up

Exposure plans for the above techniques for SRMs are established considering two aspects namely the probability of detecting

• a propellant crack of particular width and
• a gap / debond of particular width between propellant and insulation.
  
The above two in turn depend on the contrast sensitivity and resolution achievable for SRM with the X-ray machine while for the crack detection, the number of angles from which it is radiographed. The exposure plans are always such that they are aimed at the complete coverage of grain including all critical regions. A significant amount of film overlap is provided to ensure the same. For the interfacial coverage, both slit block and wire type image quality indicators (IQIs) are employed to assess the radiographic image quality. The thickness of the slit block is made 1% of the equivalent thickness of peripheral portion of the motor and slit widths ranging from 0.25 mm to 0.5 mm are used. The DIN standard wire type IQIs with wire diameters ranging from 0.5 % to 1 % of equivalent thickness are also used. For the normal coverage, step / hole type IQIs are employed.

A solid propellant motor typically consists of a number of segments. These segments are generally processed one at a time. Each of these segments are then subjected to X-ray radiography in vertical attitude on a rotary table. As per the exposure plan, various regions of each segment are radiographed by means of rotating it on a rotary table and hoisting / lowering the X-ray machine. The propellant mass is inspected by keeping the film inside the port of the segment. The X-ray beam traverses the propellant web after passing through the case and insulation. The image of defects such as voids in the propellant which lie in lines are therefore superimposed on to the same locations on a two dimensional radiograph. Thus, the information that will be obtained about the void distribution is one of axial rather than radial. As the motors are of radial burning type with the flame starting from the port, the information of radial distribution of voids assumes major significance from the ballistic point of view. These voids have the potential of augmenting the burn-rate of the motor depending on their intensity. Therefore, the estimation of radial extent of void distribution is essential to study the thermal, mechanical and structural implications. In the absence of suitable computed tomographic scanners, this is a formidable task indeed. In order to meet this requirement, specialized techniques/methods have been developed in order not only to map and locate void zones but also to estimate thickness loss due to propellant porosity / void cluster in the direction of flame front propagation. Accordingly a modified triangulation technique for locating void zones and densitometric method involving the use of micro-densitometer for estimating the third dimension of defects / the thickness loss in the depth direction have been developed in-house.

3.0 Modified Triangulation Analysis

The principle is based on imaging the shift of a defect region as the direction of X-ray beam through the object is varied from the normal direction. This involves keeping the image plane as constant and directing the X-ray beams through the object at three different angles and generating the corresponding radiographs. The first radiograph is taken by passing X-rays normal to the image plane and other two radiographs are produced by directing the X-ray beam at an angle to the normal direction, one above and other below respectively as shown in fig.2.

Fig 2: Illustration of modified Triangulation Principle

This involves either hoisting or lowering of the X-ray machine and tilting it to get to the desired orientation. The angles at which the latter two radiographs are to be taken with respect to the normal are decided by size and extent of defect region in the depth dimension of the object. The idea is to get a clear-cut separation of the images of defect region from that of normal exposure. The shift of the boundaries of the defect region with respect to that of normal exposure is measured from the corresponding radiographs. These values are then plotted on to a graph sheet. The intersection of boundaries of the defect region due to all the three exposures in the graph gives the location and size of the defect region.

To carry out the triangulation analysis, the rocket segment has to be placed on a rotary table and arrangements have to be made for locating X-films suitably inside the port. To identify the shift of the defect region, a scale fixed with lead numerals is positioned along the entire length of the port (image plane). To identify the axial extent of void zone, a normal exposure is taken by directing the X-ray beam perpendicular to the grain geometry. Once the axial distribution of voids is identified, two angular exposures are taken as mentioned above by suitably orienting the X-ray beam in each case. The deeper the defect region lies, greater should be the angle with which the X-ray beam is to be directed in order to ensure separation of the defect images from that of normal exposure. More than one film may be needed for each of the angular exposures to completely cover the shift of the defect region. The radiographs are subjected to detailed measurements on a suitable illuminator. The graphical visualization of the defect region then involves four steps. The first two steps involve

• tracing the vertical slice of whole of the segment to the scale on a graph sheet and
• tracing to the scale, in sequence, the boundaries of X-ray beams from the focal spot of the source to the image plane. The subsequent steps involve
• figuring out the intersection of the three X-ray beam boundaries on the image plane and
• reading the values of both the radial extent and location of defect regions.
The accuracy of localization of defect region improves when the region lies closer towards the periphery side rather than port side. This is due to the fact that obtaining a reasonable image shift is difficult when the defect region is close to the port as the X-ray machine maneuverability is constrained.

4.0 Densitometric analysis

It is well known that an X-ray image is the result of differences in the attenuation of X-rays between different parts of an object. These attenuation differences arise due to the interactions of X-rays with the object mainly through three major mechanisms such as

• photoelectric effect,
• Compton effect, and
• pair production.
The effect of these interaction mechanisms is such that the resulting attenuation (scattering and absorption) in general increases as a function of specimen thickness and density. The minimum thickness change (Dx) of the object that can be detected in a radiograph depends on a number of parameters such as specific subject contrast, film contrast, total unsharpness, viewing conditions etc. The mathematical expression for D x is

D x =2.303 * D D * HVL * B / (0.693 * G)

(1)

where,		DD		=		change in optical density
		HVL		=		half value layer of material
		B		=		build-up factor
and		G		=		film gradient

The build-up factor B is expressed as (1 + I_S/ I_D) and is a measure of the ratio of scattered X-rays due to the object (I_S) and primary radiation (I_D) reaching the X-ray film. B is to be determined experimentally for every radiographic condition involving different portions of SRM.

4.1 Experimental determination of B for SRM
Given the radiographic conditions, each segment of a large booster presents in general, three different regions namely two ends and one cylindrical portion. For the given X-ray machine (15 MeV linac), the scatter build-up due to these regions vary significantly due to the variation in the configuration and thickness of casing material, insulation, and solid propellant. Therefore, estimating the effective thickness change or loss (D x) due to the presence of cluster of voids in the given region of rocket motor, requires the determination of the corresponding build-up factor. B has been determined in the present case, by keeping on each portion of the segment motor, a propellant slab with introduced thickness change in the form of a slot having a depth of 30 mm. Radiographic technique used is the same as that employed under normal working conditions. The optical density change in the radiograph due to the slot is measured using a microdensitometer. The value for the known thickness change D x is then substituted in the following equation (obtained by rearranging equation [1]) to get the desired build-up factor value for the region concerned.


B = 0.3* G * D x / (D D*HVL)

(2)

The same process is repeated for slots of 5mm, 10mm, 15 mm, 20mm, and 25 mm depth. As the radiographic technique parameters are kept constant for all the above cases, the value of B remains the same as that for 30 mm slot as long as the variation of G is very small over the range of optical densities [D D] for the selected slots.

4.2 Estimation of thickness loss due to void cluster
Now having determined B, the effective thickness loss D x due to void cluster can be estimated using equation (1). For this we require the accurate determination of optical density difference D D (= D₂ - D₁), between the region of void cluster at maximum density (D₂) and the defect-free region of the propellant (D₁). Microdensitometer operating at an optimised slit width of 78 m m has been used to scan the radiographs corresponding to void clusters along straight lines and simultaneously generate optical density profiles on graph sheets. In doing so, it must be ensured that the base line (D₁) for each such profile corresponds to the region of defect-free grain. In addition, on each graph sheet, the optical density profile of a calibrated wedge should also be recorded in order to identify the optical density level at which one operates and hence appropriately take the corresponding G for the estimation of thickness loss. A typical microdensitometer trace of a void region is shown in fig 3.

Fig 3: Typical microdensitometer trace of a defect region

5.0 Conclusion

The specialized Radiographic techniques such as
• the modified triangulation analysis and
• Densitometric analysis developed in-house at SHAR Centre helped not only to estimate the radial extent of void cluster but also sizing, whenever defects are encountered.
In the absence of CT scan, the methods described, give fairly reasonable estimate of defect characterization. This data is very important to the designers in order to judiciously decide the acceptance of the expensive rocket motor for its end use.

© AIPnD , created by NDT.net

|Home|    |Top|