TABLE OF CONTENTS

Introduction Selection Of Testing Method Types of inhomogeneity Characteristics of testing methods Extent Of Testing Practice in Slovenia Influence of quality assurance (QA) and quality control (QC) Proposal for determination of the extent of testing Conclusion Literature

The 4th International Conference of Slovenian Society for Nondestructive Testing "Application of Contemporary Nondestructive Testing in Engineering" 24 - 25 April 1997, Ljubljana, Slovenia.
Table of Contents

1 INTRODUCTION

Each testing of welded joints should start with the selection of a suitable testing method and determination of the extent of testing. One should be aware that the selection of a testing method and determination of the extent of testing are as important as testing itself when the final result is considered. Owing to the selection of an unsuitable testing method NDT methods often become purposeless.

When planning NDT testing one should take into account a number of factors such as:

• requirements regarding reliable and safe operation,
• quality assurance level achieved,
• characteristics of welding processes, 'I properties of materials used,
• feasibility of NDT methods available,
• economic criteria.

In the selection of a suitable NDT method, welding technologists and NDT experts should be involved, and in determining the extent of testing designers, who know best which are the intended operating conditions of a device to be tested and which locations may be loaded most.

2 SELECTION OF TESTING METHOD

The selection of the NDT method is usually the concern of NDT laboratories, which know the NDT methods and their feasibility in detecting defects. Regarding properties of materials and characteristics of a welding process it should be first established what types of defects may occur and where. It should also be established which part of a structure may incur the highest stresses.

2.1 Types of inhomogeneity

An expert may, with a high level of certainty, predict which types of defects may occur and at which locations are they most probable. Attention should be paid to bonding defects and cracks, which are in no case admissible. A little more toleration may be shown to pores and non-metal inclusions. But a careful examination of heavily loaded parts of a structure should be carried out.

Hot cracks, i.e. crystallization cracks, usually occur in the middle of the weld. Hotcracking sensibility is found in alloys, alloyed steels and steels containing a higher level of impurities, usually in welds of larger dimensions. Cold cracks occur mostly in the heat affected zone. They occur in steels susceptible to through-hardening and hydrogen embrittlement.

A special attention should be paid to porosity in gas-shielded welding, submerged arc welding, and welding with basic electrodes.

Typical cracks in the welded joint and location of their appearance is shown in Fig. 1.

Characteristic location of cracks Fig. 1. Typical weld defects

2.2 Characteristics of testing methods

As to their ability of detecting various geometrical forms of defects, the NDT methods applied in testing of welded joints differ one from the other very much, i.e. they complement each other (1, 2, 3). In some cases they are interchangeable (4).

In testing of welded joints, it is radiographic methods which are most frequently used and which permit a very reliable detection of three-dimensional discontinuities such as pores, non-metallic inclusions, incomplete penetration and undercuts at the inaccessible root side. The method seems to be less reliable in detecting planar, i.e. two-dimensional, defects such as cracks.

The ultrasonic methods seem to be the most universally applicable. They may be applied to all types of defects but they are comparatively complicated and sensitive to various disturbances. They are less reliable, therefore, they are making themselves valued in welding very slowly.

Simple and reliable methods are available for detection of cracks reaching the surface. Magnetic methods are suitable for ferromagnetic materials, while penetrant methods are suitable for all metals.

Too little attention is, however, paid to visual inspection which should be performed prior to each NDT examination. The visual inspection provides basic information on the state of welded joints and the structure concerned. This is a guideline for further examinations. Fig. 2 schematically shows the application of various NDT methods in testing butt and fillet welds.

Fig. 2. Application of NDT methods

In addition to technical considerations, in the selection of testing methods the cost of testing should be considered as well. The more perfect weld is to be obtained, the more should be invested into NDT methods (Fig. 3). Fig. 4 shows market rates for NDT methods in Slovenia.

Fig. 3. Ratio between control cost and weld perfection	Fig. 4. Standard market rates for NDT methods in Slovenia

3 EXTENT OF TESTING

The extent of testing is a datum informing us what part of the total length of welded joints and which locations is to be tested. The extent of testing is sometimes determined by the client in the invitation to project tender, but most frequently by a designer taking into account technical regulations in force and stresses calculated for individual welded joints.

3.1 Practice in Slovenia

In the techical regulations in force provisions on the minimum extent of NDT testing of welded joints may be found. Although some are rather old they are still in force (5, 6). They are not consistent with each other and they may be applied only to certain fields. An exception is the standard on quality levels with a general applicability (7).

For steam boilers and steam superheaters a 30 % radiographic examination of butt welds is specified (5).

With welded structures the extent of testing depends on the welding location and the joint quality class selected (6). Welds made in a workshop require less testing than out-of-position field welds. Regarding the load and the level of requirements, three classes are distinguished: special, first and second.

The regulations on pressure vessels are the most elaborate ones. The extent of testing depends on operating conditions of the vessel (vessel class), the type of joint and strength calculation (8, 9, 10). There are four classes of welded joints. A selection may be made among various levels of perfection of welded joints.

A similar division into four quality classes may be found in the general quality standard (7) which is applied to pipelines, with the exception of main pipelines for liquid and gas hydrocarbons where the extent of testing is regulated in a different way (11). The latter depends on the distance from settlements, energy facilities and traffic routes.

Among the methods for testing welds, the radiographic method is preferred. In older regulations other methods are not mentioned. More recent regulations permit application of ultrasonic methods instead of radiography. Other methods are applied only as a complement.

Table I gives a survey of obligatory application of NDT methods in Slovenia.

3.2 Influence of quality assurance (QA) and quality control (QC)

When selecting the extent of testing, the level of quality assurance achieved should be taken into account. This has not been taken into account in Slovenia till now. Also some foreign regulations known in Slovenia do not take quality assurance into account.

Fig. 5. Ratio between quality assurance (QA) and quality control (QC)

More recent standards specify three levels of quality assurance, which permits, when selecting the extent of testing, to give consideration also to the level of quality assurance achieved by the performer of welding work. The welded-joint quality (Q) may be defined as a sum of quality assurance (QA) and quality control (QC).

The maximum quality in the instance selected may be achieved if all means of quality assurance available are taken into account and all NDT methods available are applied. The relation between quality assurance (QA) and the extent of quality control (QC) is graphically shown in Fig. 5.

Table I: Survey of testing methods specified
Field of application	Level of requirements	Extent and method of testing
Steam boilers and steam superheaters	-	30 % RT
Steel structures	S I II	100 % RT, MT and PT if required 10-50 % RT, MT and PT if required RT of out-of-position welds
Pipelines JUS C.T3.010	I II III IV	100% RT or VT, 100 % MT or PT 50 % RT or VT, 100 % MT or PT 10 % RT or VT, 30 % MT or PT -
Pressure vessels	I II A II B III A III B IV A IV B I c II c I D II D III D	100 % RT or VT, 100 % MT or PT 70 % RT or VT, 100 % MT or PT 50 % RT or VT, 100 % MT or PT 30 % RT or VT, 30 % MT or PT 10 % RT or VT, 30 % MT or PT no requirements 30 % MT or PT 100 % UT, 100 % MT or PT 50 % UT, 50 % MT or PT 100 % UT, 100 % MT or PT 70 % UT, 100 % MT or PT 310%UT, 30%MT or PT
Main pipelines	Zone I Zone II Zone III	10 % RT, MT or PT if required 50 % RT, MT or PT if required 100 % RT, MT or PT if required

- For all quality levels a 100 % visual inspection is required.
- MT and PT are performed only if there is a risk of cracks.

Table II: Extent of testing as a function of QA and requirements
QA level	Requirements
	Low	Medium	High
III	0%	25%	50%
II	0%	50%	75%
I	25%	75%	100%

Quality assurance requires a number of operations usually expressed by levels. The new European standard specifies elementary (1), standard (11) and comprehensive (Ill) levels (12).

The level selected may be achieved by a higher level of quality assurance and less control (instance a) and vice versa (instance b). Both instances are shown in Fig. 6.

Similar conclusions may be drawn if quality assurance and quality control are considered production costs and both together quality cost. In this case the following holds true:

Q = QA + QC (monetary unit)

Fig. 7 shows a cost model for quality.

Fig. 6. Quality selection between QA and QC	Fig. 7. Cost model for quality

It may be concluded that quality assurance and control complement each other and jointly produce quality. In practice both means, however, are not balanced, which is shown by the model (Fig. 7). In certain cases quality assurance is a more efficient means than quality control and vice versa; therefore, each is used as appropriate.

In practical applications a certain lower threshold is more and more being taken into account; these are a welder's approval test certificate in quality assurance and the visual examination in quality control. For pressure vessels and load-bearing welded structures the lower threshold is determined by the requirements related to the issue of a certificate of qualification to the workshop concerned.

3.3 Proposal for determination of the extent of testing

When determining the extent of testing, beside requirements for reliable and safe operation of the structure, also the level of quality assurance already achieved by the performer of welding work should be considered. Structures may be divided into three classes, i.e. very exacting, exacting, and less exacting.

The ratio between the level of quality assurance achieved, the requirements, and the extent of testing may be graphically presented (Fig. 8). Table 11 shows how the extent of testing may be determined as a function of the level of quality assurance achieved.

We have been used to standards determining quality classes as well as the minimum extent of testing. Recent standards have no such provisions because the level of acceptance of defects, which is determined by their size and number, has nothing to do with the extent of testing. But in any case a higher level indicates a higher quality of the welded joint (Fig. 9). Quality levels B, C, and D were taken from the new European standard (13).

Fig. 8. Extent of testing as function of QA and requirements	Fig. 9. Schematic representation of quality levels of welded joints

4 CONCLUSIONS

Quality assurance and quality control complement each other. With a higher level of quality assurance the extent of testing may be reduced. The above-mentioned principle would certainly stimulate performers of welding work to invest more into quality assurance.

It is suggested that the above-mentioned principle of determining the extent of testing be implemented in the elaboration of new Slovenian technical regulations.

5 REFERENCES

1. IlW Guidance on Assessment of the Fitness for Purpose of Welded Structures. Doc. IIS/IlW-SST-1 157-90 (pp. 67-74, 229-296).
2. P. Klug: Quality Management in the Field of Welding.
3. H. Granjon: Metalurske osnove varjenja, Zveza drustev za varilno, tehniko Slovenije, Ljubljana 1994 (pp. 238-241).
4. Guidelines for Replacing NDE Techniques with one another. IIS/IIW Doc. V- 1062-96.
5. Pravilnik o tehnicnih predpisih za izdelavo in uporabo parnih kotlov, parnih posod, pregrevalcev pare in ogrevalcev vode. Ur. 1. SFRJ st. 7-157/57.
6. Pravilnik o tehnicnih pogojih in normativih za varen transport tekocih in plinastih ogljikovodikov, clen 61. Ur. 1. SFRJ st. 26/85.
7. JUS C.T3.010: Zavarivanje i srodni postupci. Klase kvaliteta zavarenih spojeva izvedenih topljenjem na celiku.
8. Pravilnik o tehnicnih normativih za stabilne tlacne posode. Ur. 1. SFRJ st. 16/83.
9. Pravilnik o tehnicnih normativih za premicne zaprte posode za komprimirane, utekocinjene in pod tlakom raztopljene pline. Ur. 1. SFRJ st. 25/80 in 9/86.
10. JUS M.E2.159: Posude pod pritiskom. Kontrola i ispitivanje zavarenih spojeva.
11. Tehnicni predpisi o kvaliteti zvarnih spojev za nosilne jeklene konstrukcije. Ur. 1. SFRJ 19/64.
12. SIST EN 729: Zahteve po kakovosti pri varenju - Talilno varjenje kovinskih materialov
    1. del: Smernice za izbiro in uporabo
    2. del: Obsimejse zahteve po kakovosti
    3. del: Standardne zahteve po kakovosti
    4. del: Osnovne zahteve po kakovosti.
    
13. SIST EN 25817: Oblocni zvarni spoji na jeklu - Smernice za stopnje sprejemljivosti napak.