Method to determine preheat temperature of ferritic steels

The methods, described in EN 1011-2:2001, are recommendations to avoid hydrogen

cracking (also known as cold cracking) in ferritic steels.

Many methods have been proposed for predicting preheat temperature to avoid hydrogen
cracking in non-alloyed, fine grained and low alloy steel weldments.
Examples are given in IIW documents.
Two of those methods are described in this standard:

Method A is based on extensive experience and data which is mainly, but not exclusively, for
carbon-manganese type steels.
Method B is based on experience and data which is mainly, but not exclusively for low alloy
high strength steels.

Beside these two methods there are tables which shall be used for creep resisting and low
temperature steels. (method C).

The recommendations apply only to normal fabrication restraint condition.

Higher restraint situations may need higher preheat temperature or other precautions to
prevent hydrogen cracking.
The methods A and B refer to welding of parent metal at temperatures above 0 °C. When
welding is carried out below this temperature it is possible that special requirements will be
needed.
Otherwise lower preheat temperatures are possible, if this is supported by procedures.

To calculate the preheat temperature for method A or B you have to know

 The hydrogen content of the consumable (HD)
 The composition of the parent metal (CE of CET);
 The plate thickness and joint geometry
 The heat input

In this program, a method D is added, based on the standard AWS D1.1

To calculate the preheat temperature for this method you have to know
 The hydrogen content of the consumable (HD)
 The composition of the parent metal (Pcm);
 The plate thickness
 The restraint level
Weld temperature cycle

The calculation of the welding temperature cycle is based upon the simplified formulas of
Rykalin. The formulas used here for respectively 3- and 2-dimensional cooling of a bead on
plate are the following.

3-Dimensional:

The temperature as a function of time and place is given by

Q  R2 
T(t, R )   exp    T0 .
2  t  4at 

The cooling time from 800°C to 500°C then is

Q  1 1 
t 8 / 5     .
2   500  T0 800  T0 

2-Dimensional:

The temperature as a function of time and place is given by

Q  R2 
T(t, R )   exp    T0 .
d 4 c t  4at 

The cooling time from 800°C to 500°C then is

Q2  1 1 
t 8 / 5  2
  2
 .

4 cd  ( 500  T0 ) ( 800  T0 ) 2 

Here R represents the distance to the center of a point (3D) or line (2D) shaped heat source,
 , c en  are the physical “constants”, d is the plate thickness and T 0 the preheat and
interpass temperature.

The relations for the cooling time t8/5 have been empirically adapted to steel by Uwer et. al.
(IIW doc. IX 1631-91), obtaining the formulas below. Here there is no need for the values of
,  en c, which are often difficult to obtain. Furthermore the weld shape factor for three- or
two-dimensional heat flow (F3 respectively F2) has been introduced. This enables one to
calculate more situations than a bead on plate. These new formulas for t8/5 are described in
EN 1011-2.
3-Dimensional:
 1 1 
t 8 / 5  (6700  5 T0 )  Q      F3 .
 500  T0 800  T0 

2-Dimensional:
Q2  1 1 
t 8 / 5  (4300  4, 3 T0 )  10 5  
 ( 500  T )    F2 .
d2 (800  T0 ) 
2 2
 0

The transition thickness dt is the plate thickness at which the transition from three-
dimensional to two-dimensional heat flow takes place. In that case F2 = F3 and both values
of t8/5 are equal, also:

( 4300  4,3  T0 )  10 5  1 1 
dt   Q    
6700  5  T0  500  T0 800  T0 

Some values of the transition thickness are below:

Preheating temperature
Q.
20°C. 100°C. 200°C.
0,5 10.4 11.1 12.3
1 14.7 15.7 17.4
1,5 18 19.2 21.3
2 20.7 22.2 24.6
2,5 23.2 24.8 27.5
3 25.4 27.1 30.1