Impact Testing of Welds

The impact testing of welds and some of the factors that affect the transition temperature such as composition and microstructure. Within such a short article, however, it will only be possible to talk in the most general of terms.

Welding can have a profound effect on the properties of the parent metal and there may be many options on process selection, welding parameters and consumable choice that will affect impact strength.

Fig.1. PD5500 App D. location of
Charpy specimens in weld HAZ

Many application standards therefore require impact testing to be carried out on the parent metal, the weld metal and in the heat affected zone as illustrated in Fig.1 which is taken from BS PD 5500 Annex D. The standards generally specify a minimum impact energy to be achieved at the minimum design temperature and to identify from where the specimens are to be taken. This is done in order to quantify the impact energy of the different microstructures in the weld metal and the HAZs to ensure that, as far as possible, the equipment will be operating at upper shelf temperatures where brittle fracture is not a risk.

These application standards may be supplemented by client specifications that impose additional and more stringent testing requirements, as shown in Fig.2 taken from an oil industry specification for offshore structures.

Fig.2. Offshore client requirements

The positioning of the specimens within a weld is extremely important both in terms of the specimen location and the notch orientation. A specimen positioned across the width of a multi-pass arc weld will probably include more than one weld pass and its associated HAZs. Quite a small movement in the position of the notch can therefore have a significant effect on the impact values recorded during a test. Positioning a notch precisely down the centre line of a single pass of a submerged arc weld can give extremely low impact values!

Testing the heat affected zone also has problems of notch position since in a carbon or low alloy steel there will be a range of microstructures from the fusion line to the unaffected parent metal. Many welds also use a 'V' preparation as illustrated above and this, coupled with the narrow HAZ, means that a single notch may sample all of these structures. If the impact properties of specific areas in the HAZ need to be determined then a 'K' or single bevel preparation may be used.

The standard specimen is 10mm x 10mm square - when a weld joint is thicker than 10mm the machining of a standard size specimen is possible. When the thickness is less than this and impact testing is required it becomes necessary to use sub-size specimens.

Many specifications permit the use of 10mm x 7.5mm, 5mm and 2.5mm thickness (notch length) specimens. There is not a simple relationship between a 10mm x 10mm specimen and the sub-size specimens - a 10mm x 5mm specimen does not have half the notch toughness of the full size test piece. As the thickness decreases the transition temperature also decreases, as does the upper shelf value, illustrated in Fig.3 and this is recognised in the application standards.

Fig.3. Effect of size on transition temperature
and upper shelf values

In a carbon or low alloy steel the lowest impact values are generally to be found close to the fusion line where grain growth has taken place.

Coarse grains generally have low notch toughness, one reason why heat input needs to be controlled to low levels if high notch toughness is required.

For example, EN ISO 15614 Pt. 1 requires Charpy-V specimens to be taken from the high heat input area of a procedure qualification test piece and places limits on any increase in heat input. Certain steels may also have an area some distance from the fusion line that may be embrittled so some specifications require impact tests at a distance of 5mm from the fusion line.

Charpy-V tests carried out on rolled products show that there is a difference in impact values if the specimens are taken parallel or transverse to the rolling direction. Specimens taken parallel to the rolling direction test the metal across the 'grain' of the steel and have higher notch toughness than the transverse specimens - one reason why pressure vessel plates are rolled into cylinders with the rolling direction oriented in the hoop direction.

In a carbon or low alloy steel the element that causes the largest change in notch toughness is carbon with the transition temperature being raised by around 14°C for every 0.1% increase in carbon content.

An example of how this can affect properties is the root pass of a single sided weld. This often has lower notch toughness than the bulk of the weld as it has a larger amount of parent metal melted into it - most parent metals have higher carbon content than the filler metal and the root pass therefore has a higher carbon content than the bulk of the weld.

Sulphur and phosphorus are two other elements that both reduce notch toughness, one reason why steel producers have been working hard to reduce these elements to as low a level as possible. It is not uncommon for a good quality modern steel to have a sulphur content less than 0.005%.

Of the beneficial elements, manganese and nickel are possibly the two most significant, the nickel alloy steels forming a family of cryogenic steels with the 9% nickel steel being capable of use at temperatures down to -196°C. Aluminium is also beneficial at around 0.02% where it has the optimum effect in providing a fine grain size.

Lastly, let us have a brief look at some of the other factors that can affect the impact values. These are concerned with the quality of the specimen and how the test is conducted.

It goes without saying that the specimens must be accurately machined, the shape of the tip of the notch being the most important feature. A blunted milling cutter or broach will give a rounded notch tip and this in turn will give a false, high impact value. Checking the tip radius on a shadowgraph is one simple way of ensuring the correct tip shape. Correct positioning of the specimen on the anvil is most important and this can be done using a specially designed former.

The last point concerns the testing of specimens at temperatures other than at room temperature. When testing at sub-zero temperatures the length of time taken to remove the specimen from the cooling bath, position it on the anvil and test it is most important. EN875 requires this to be done within five seconds otherwise the test piece temperature will rise making the test invalid - referring back to the impact energy vs temperature curve in the previous post will show why.

Relevant Specifications
BS 131	Part 4	Calibration of Impact Testing Machines for metals.
BS 131	Part 5	Determination of Crystallinity
BS 131	Part 6	Method for Precision Determination of Charpy-V Impact Energy
BS 131	Part 7	Specification for Verification of Precision Test Machines
EN 875		Destructive Tests on Welds in Metallic Materials - Impact Tests
EN 10045	Part 1	Test Method
EN 10045	Part 2	Verification of Impact Testing Machines
ASTM E23-O2A		Standard Test Methods for Notched Bar Impact Testing of Metallic Materials.