By Spencer Erling,
Education Director, SAISC
WHAT IS HYDROGEN EMBRITTLEMENT?
We used the term a number of times in part 1 without defining the term. Thanks to Wikipedia for the following definition: Hydrogen embrittlement is the process by which various metals, most importantly high-strength steel, become brittle (and or a decrease of toughness or ductility) and fracture following exposure to hydrogen. Hydrogen embrittlement is often the result of unintentional introduction of hydrogen into susceptible metals during forming (molten state) or finishing operations (elevated temperature treatments and could include hot dip galvanizing or electroplating), together with high stress levels (which could be residual stresses from manufacturing processes or induced stress such as stretching during the tightening of bolts or increase in loads in the structure) and (increases) cracking in the material. This phenomenon was first described in 1875.
The mechanism starts with lone hydrogen atoms diffusing through the metal.
In broad terms hydrogen embrittlement can be prevented through:
1. Control of stress levels (not possible with 14399 bolts due to the very nature of the pre-loading process)
2. Lower levels of hardness (ditto)
3. Avoidance of the hydrogen source(s)
4. Making the finished product to remove hydrogen
CAN CHEMICAL COMPOSITION AFFECT HYDROGEN EMBRITTLEMENT?
Whilst SANS EN 898 part 1 does define in broad terms chemical composition of the bolts, the compositions listed are sometimes in ranges of percentages (i.e. between minimum and maximum). So whilst a steel maker could well conform with the ranges, there are schools of thought that are suggesting that having the maximum percentage of one or more alloys could make the steel more susceptible to hydrogen embrittlement. It appears as if a lot more research is required in this regard.
WHAT IS THE INFLUENCE OF THE INSTALLER OF THE BOLTS?
There is nothing specifically documented in this regard. What we know is that some time after installing and apparently correct tightening of bolts we have had bolts snapping (admittedly a minute percentage in any one project of the bolts do actually fail).
An example of what could cause failure is abuse by the installer. Imagine an end plate to a large (deep) plate girder. It is common knowledge that end plates are never totally flat after welding and handling. By definition, for our standard range of South African beams, end plates are described as flexible end plates to allow for rotation of the ‘simply supported’ beam. So they are usually designed as relatively thin material (8mm plates).
However for plate girders it is common practice to use thicker plates. When they are installed, by definition the end plates should pull up flat with the component it is attaching to. When they do not pull up flat erectors often use the connection bolts to do this flattening work (which really should have been done before erecting). It is anyone’s guess how much effort goes into pulling up the plates to be in contact and what stress they have induced.
If the turn-of-the-nut-method is used it is very likely that tightening is way beyond snug tight (by definition in SABS10094 as 10% of proof stress load). The additional part turn therefor stresses the bolt way above the intended tension, leading to who knows what happens when live loads come on to the structure.
The least a contractor should do in this circumstance is to replace the bolts he has used to pull the plates up to flat tight contact with new bolts which should then be tightened using the correct procedure.
I sincerely hope that bad procedures on site have not been responsible for the unbelievably expensive, both in time and money terms, for those projects affected so far.
WHAT ABOUT HOLDING DOWN BOLTS?
The current version of the Red Book (The SA Steel Construction handbook) recommends that when HD bolts are required to conform to 8.8 or 10.9 mechanical requirements, EN19 material should be used. After machining, the HD bolt should be heat treated to the correct hardness value to ensure a suitable ultimate tensile strength equivalent to 8.8 and 10.9.
Whilst the instruction is sound, what we did not warn users about was that it is critical not to weld (even tack weld) such hardened bolts unless using a properly prepared prequalified weld procedure (as envisaged by the AWS D 1.1 welding specification), with a lot of emphasis on the pre and post heating of weldments.
There have in the last two years been two HD bolt failures directly related to no preheating of weld areas (including a tack weld in once instance). In both cases the bolts were hot dip galvanized. We do not know if the process (in particular the pickling process which is dipping in acid) exacerbated the problem.
The SAISC now recommends the following for HD bolts.
1. For most static structural applications use commercial quality round bar (reinforcing bar) up to the loads at which bolt diameters do not exceed 30mm (200 MPa yield).
2. For diameters exceeding 30mm, graded material (S355JR) is available.
3. These two materials are easily machined, welded and hot dip galvanized.
4. For dynamic situations the grade 8.8 or 10.9 equivalent described above can be used and pre-tightened to an agreed specification.
5. We strongly recommend eliminating of welding to these materials if possible by:
a) Instead of welding a pull-out plate at the bottom of the bolt for casting into concrete, we suggest cutting a thread and bolting the pull-out plate between two nuts.
b) We also suggest that templates be attached by bolting once again with two nuts, one of which will be under the template and one above.
c) All of these loose components can be galvanized taking the same care and processes one would use for grade 10.9 bolts.
There is no doubt that hot dip galvanizing is still a great value for money corrosion protector in the correct environment. For grades 4.8 and 8.8 bolts as stated above there are no issues relating to the galvanizing of these bolts.
However, readers should ensure that when they are purchasing grade 10.9 bolts in hot dip galvanized finish that they clearly spell out the standards applicable i.e. ISO898 and EN 14399. They should purchase bolts from reputable suppliers and ensure that the manufacturer’s fully traceable certification accompanies the bolts at all times.