Cold cracking and weldability testing

Depending on the chemical composition of the steel, of the design of the joint to
welding and the amount of hydrogen introduced into the welding area (by
the ambient humidity for example), the welds can be more or less
sensitive to cold cracking. This type of cracking appears
generally at room temperature and within about 48 hours afterwards
the end of the welding operation, even in the absence of a load
exterior. In order to prevent the risk of cold cracking, a very large
number of tests exist, some of which are standardized in the
NF EN ISO 17642 (Part 2 and Part 3). In this Métalétech publication,
three of the main types of cold cracking tests are presented and their
the respective field of application is explained.

Generalities

First of all, it is important to clarify that the following tests should not
to be confused with the qualification methods of welding procedures
(QMOS) referenced in the NF EN 1090-2. Indeed, the objective of QMOS is to qualify.
a welding procedure (i.e. to demonstrate that the chosen parameters are
adapted for the planned welding operation) which will be used in production by the
continue with a description of the operating procedure derived from this QMOS. This qualification can
to be carried out using tests (for example with pre-production welded assemblies
according to NF EN ISO 15613) or without specific tests (for example using a mode
standard welding procedure – see NF EN ISO 15612). The choice of the method of
qualification depends notably on the complexity of the assembly and the execution class
(see NF EN 1090-2 for more details).

Weldability tests, for their part, are conducted in order to study feasibility.
of a welding operation in terms of welding parameters, base materials
used (for example, assembly of different materials) and pre-/post-heating as well
in terms of optimizing these parameters. The weldability tests can then be
relevant for preparing QMOS in certain situations. It can be noted that it is not
generally not necessary to carry out specific weldability tests when the
standard construction steel plates or profiles (for example S235 to S355) are used
and if the thicknesses do not exceed approximately 70-80 mm. On the other hand, for high-strength steels
and very high yield strength (for example S690), for significant thicknesses and, in
in certain conditions, for stainless steels, the weldability tests may be
relevant in order to optimize the operating mode and prepare the QMOS. Three of the main
Weldability tests concerning cold cracking are then presented below.
Before presenting them, it seems interesting to recall some key elements.
concerning the phenomenon of cold cracking.

Cold cracking

We can distinguish three main types of cold cracking:

• Cracking due to material embrittlement/hardening;

• Hydrogen cracking;
• Laminated tearing.

André Beyer – Research Project Manager – CTICM – August 2020 1

Hardening cracking is primarily caused by the formation of martensite.
in too large a quantity in the welding zone and the thermally affected zone. The
martensite, being a metallurgical phase (crystal lattice configuration) is brittle, reduces
locally the ductility of steel.

The sensitivity of construction steels to the formation of martensite can be

estimated using the carbon equivalent value CEV:

CEV = C + Mn/6 + (Cr+Mo+V)/5 + (Ni+Cu)/15

where

C: What is the percentage of Carbon in the alloy?

Mn: is the percentage of Manganese in the alloy
Mo: is the percentage of Molybdenum in the alloy
V: What is the percentage of Vanadium in the alloy?
You: What is the percentage of Nickel in the alloy?
Cu: is the percentage of Copper in the alloy

Generally, steels with a CEV lower than 0.4 are less sensitive to embrittlement.
under normal conditions (for example: low hydrogen input) while the
Alloys with a CEV greater than 0.7 are sensitive to embrittlement and cracking at
cold. By carrying out one of the tests presented below, it is still possible to
déterminer des paramètres permettant de souder les aciers avec un CEV important.

Other than through the creation of martensite, the sensitivity to cold cracking is
influenced by the amount of hydrogen in the weld zone. Hydrogen can be
introduced into the molten metal, particularly by the humidity of the surrounding air, the shielding gas
and the added metal. Hydrogen is dissolved in the molten metal. When the temperature of the
as the metal decreases, the solubility of hydrogen also decreases. Consequently, some
hydrogen molecules re-form and diffuse particularly in areas of
martensite. Due to the accumulation of hydrogen molecules, the pressure increases.
considerably in the crystal network which causes cracking. In order to prevent the
hydrogen-induced cracking, it is for example possible to use some
additive materials with a controlled and sufficiently low hydrogen content.

Finally, lamellar tearing is, under certain conditions, caused in the area
thermally affected. Indeed, rolling can create surface inclusions of
manganese disulfide which reduces the properties in the thickness direction of the steel
of base. To avoid lamellar tearing, it is recommended to choose qualities Z
steel according to NF EN 1993-1-10 (Eurocode 3 Part 1-10).

It can be noted that cold cracking tests primarily aim at hydrogen cracking.
and the cracking due to embrittlement, which are strongly related to each other. As mentioned, it
There are possibilities to prevent these two types of cracking through adequate measures.
simple as the choice of filling materials and the pre- and/or post-heating that the
temperature can be chosen based on the CEV (or other calculation formulas of a
equivalent carbon). However, for certain steels (very high strength construction steels)
elastic limit, certain stainless steels), these simple approaches are not
sufficient or lead to difficult welding conditions (for example, pre- and post-
important heating). Therefore, in order to find the welding parameters and to
Optimizing the implementation of weldability tests can be wise. These tests are quite
simple and create fairly realistic throttling effects.

André Beyer - Research Project Manager - CTICM - August 2020 2

Self-limitedTestTekken

The self-limited trial Tekken (the name comes from the Japanese Tesudo Gijutsu Kenyushi) has been
developed in the 1960s to determine sensitivity to cracking
cold of the thermally affected zone and the weld zone (melted zone). This test
is standardized in the NF EN ISO 17642-2 standard. In the standardized version, the test is
carried out with a Y bevel, but in order to reproduce comparable conditions to the
real welding to be made, other bevel preparations can be used. The
Figure 1 shows a Tekken test sample with a double V bevel. After the application
Instead of the two soldering pieces, two anchor welds are made at the ends.
of the sample (see Figure 2). These anchor welds will create the clamping during the
the execution of the test weld. The single-pass test weld is then carried out in
the central zone of the sample. Due to the constraints generated by the anchor welds,
Residual stresses resulting from the thermal cycle will affect the test weld.

Figure 1: Sample for a Tekken test

a) Setup b) Sample with anchor welds

Figure 2: Plan view of the Tekken test

The part is then stored for at least 48 hours to allow the crack to form.
After this period, a visual inspection is carried out to determine the length of the cracks.
possible surfaces. In addition, metallographic examinations are carried out in five sections.
along the weld test to check for the absence of cracks in the cut.

The Tekken test allows for testing different combinations of filler materials (y
including different hydrogen levels), pre- and post-heating conditions and
possibly preparing a seal in order to determine the optimal combination
preventing any cracking. The Tekken test is particularly used for steels.
high elasticity limit and end welds.

André Beyer - Project Research Manager - CTICM - August 2020 3

Cappedself-testCTS

Unlike the Tekken test, the CTS (Controlled Thermal Severity) test is used to
evaluate the weldability in the case of corner welds. This test, developed in the years
50 by Cottrell (engineer at the Welding Institute), is also part of the self-braced tests.
The test is conducted on a bolted plate on a second plate (support plate)
for the installation. Anchor welds are then placed on the edges
non-etched areas of the upper plate (see Figure 3). The notch, with a height of 1.6 mm and
depth 10 mm is created in order to generate a multiaxial stress field and thus
to increase the severity of the test. The corner beads of the test weld are made
in one pass. Following the completion of the first test weld, the end of the plate
the support located opposite the weld is immersed in a current of water at a
depth of about 60 mm. When the weld has reached room temperature (20°C),
The sample can be removed from the bath and must be stored for at least 48 hours before
to proceed with welding the second edge. The second fillet weld is then cooled
under the same conditions as the first and the sample is stored again for
48 hours.

Figure 3: CTS Test

Following the storage period, the sample is cut for examination.

metallographic to check for the absence of cracks.

The CTS test is primarily used for high and very high yield strength steels.
allows to determine the welding parameters such as the filler metal (and the rate
hydrogen) and the acceptable pre- and post-heating conditions to avoid the
fissuration.

Essaisurimplant

Outside of self-limited tests, there are a number of tests that introduce a

direct solicitation in welding. This solicitation represents the residual stresses
which are generated by the thermal cycle during actual welding. Generally, the aim is to

André Beyer – Research Project Manager – CTICM – August 2020 4

elastic limit of the material. The implant test, standardized in the standard
NF EN ISO 17642-3 and developed by Henry Granjon at the Welding Institute in the years
60 represents an example of a test under external loading. Figure 4 and Figure 5
schematically show this trial. The cylindrical implant with a notch.
(in circular or helical form) is inserted into a hole drilled in the support plate
which is made from the same material as the implant. Generally, an implant of a
a diameter of 8 mm is used.

The implant is inserted into the hole and, after a possible preheating, the welding wire.
is deposited in a single pass over a length of 150 mm passing over the implant. The
Weld penetration must be chosen (using appropriate welding energy)
so that the cut is located in the thermally affected area.

Figure 4: Implant Test - Plan View

Figure 5: Implant test - Section

After the completion of the welding operation, the implant is loaded with a tensile force.
when the welding area reaches a temperature calculated based on the temperature
preheating (see NF EN ISO 17642-3). The load is then maintained for 16
heures. Lors de cette phase de chargement, l’implant peut potentiellement se rompre. Si
this is not the case, cracks that may have developed in the area of
The notches can be detected by a metallographic examination.

Thanks to this test, it is possible to determine acceptable welding parameters.

to prevent cracking for a given level of applied stress. Generally, one
can consider that the effects of natural bracing in weld joints generate a level
of constraints close to the elastic limit of the base material. That is why the test

André Beyer - Research Project Manager - CTICM - August 2020 5

on an implant is often used by applying a force that creates a stress equal to the
elasticity limit at the notch.

The implant test can be used to study sensitivity to cold cracking.

from the base material and the result is transferable to a root pass.

Conclusion

This document has recalled the main origins of cold cracking.

Then, we presented some of the many usable essays for studying the
sensitivity to cold cracking of welded materials. It is still appropriate
It should be noted that the purpose of this document is to provide the main elements of these trials.
It is advisable to refer to the different parts of the standard NF EN ISO 17642 and the
relevant publications such as those published by the International Institute of Welding (IIW)
to obtain all the necessary details for carrying out the tests presented here.

References

NF EN 1090-2: Execution of steel structures and aluminum structures – Part

2 : Exigences techniques pour les structures en acier. AFNOR. Juin 2018.

NF EN ISO 15612: Description and qualification of a welding procedure for the

metal materials – Qualification by reference to a welding procedure
standard. AFNOR. July 2018.

NF EN ISO 15613: Description and qualification of a welding procedure for

metal materials – Qualification based on a welded preproduction assembly.
AFNOR. October 2004.

NF EN ISO 17642-1: Destructive testing of welds on metallic materials – Tests

Cold cracking of welded joints - Arc welding processes - Part 1:
Generalities. AFNOR. July 2005.

NF EN ISO 17642-2: Destructive testing of welds on metallic materials. Testing of

Cold cracking of welded joints – Arc welding processes – Part 2: Tests
on self-clamping test tube. AFNOR. July 2005.

NF EN ISO 17642-3: Destructive testing of welds on metallic materials - Tests

Cold cracking of welded joints – Arc welding processes – Part 3 :
Tests on a specimen subjected to an external load. AFNOR. Juliet 2005.

T. Kannengiesser, T. Boellinhaus, Cold Cracking Tests – An Overview of Present

Technologies and Applications, BAM – Federal Institute for Materials Research and Testing
and iiw,

André Beyer - Research Project Manager - CTICM - August 2020 6