Structural bolting issues in south africa

PART 1
By Spencer Erling,
Education Director, SAISC

BACKGROUND
I am sure it is common knowledge that the new Eskom power station developments are largely based on European design codes and bolting requirements. This has brought to our attention a number of issues where European bolting practice is quite different to those practices we have used in South Africa for a long time.

THE RED BOOK RECOMMENDATIONS
Chapter 6 of the Southern African Steel Construction Handbook covers the existing South African practice in reasonable depth. Issues relating to which DIN or ISO standards bolts are commonly available are quite clear.

You will notice that the mechanical properties shown in Table 6.1 did not call up the hardness requirements for the bolts (because so little emphasis is placed on this in the old versions of ISO898-1). You will read below that hardness has now become the final acceptance criteria for bolts to this standard. The general remark ‘Grade 8.8 bolts’ usually resulted in the right bolts arriving on site.

Section 6.1.1 covers ‘ordinary bolts and nuts’ used for structural applications where ‘ordinary’ means the bolts are neither pre-tensioned nor High Strength Friction Grip (HSFG) applications. We continue to advise that all structural elements should be bolted using Class Grade 8.8 bolts. Only minor elements such as purlins, girts, hand railing, stairs and the like be joined with Class Grade 4.8 bolts.

Our design tables are set up with the assumption that the shear plane is always in the threaded portion of the bolt. In the case of the shear plane being specifically in the unthreaded shank then the design engineer has to determine the value of the strength of the bolt by multiplying the bolt strength (read in the tables) by a factor of 1.43.

It is also assumed that the tightening of such bolts be done by a normal person applying as much effort as he can to a standard spanner for the bolt size in question. No washers are required for standard round 2 or 3mm oversize holes unless there is a need to protect previously applied corrosion protection.

We recommended that pre-tensioned bolts should only be used specifically where required such as in connections where the bolts are in tension (and expect reasonably high stresses in tension) or in vibrating structure applications. The principle behind this application being that when the tension in a bolt fluctuates, once pre-tensioned, the bolt will never be in a loose state that could lead to nuts falling off.

Section 6.1.2 covers HSFG bolts. For non-slip applications we recommended the use of HSFG bolts to Class Grade 10.9 (S) and tightened using the turn of the nut method. The use of a hardened washer under the turning part (usually the nut) is obligatory.

In all of the above the DIN or ISO specification determines the length of thread on the shank of the bolt unless the bolt was ordered as a set screw i.e. one with thread the whole way along the length. This is the South African definition of a set screw (see photo on page 35). The European definition for a set screw is a threaded length of rod without a head but the one end is prepared with a hexagonal indentation to allow for tightening with an Allen key as per Diagram 1 on page 35.

Because of our design approach that the shear plane will always be in the threaded portion of the bolts, set screws (as we define them bolts threaded for the full length of the shank) are perfectly acceptable.

Where for some particular reason the designer chose to insist that the shear plane would be in the unthreaded shank, getting the correct length of bolt to ensure the shear plane will always be in the unthreaded shank could in some instances have been a problem for the unwary buyer who just ordered grade 8.8 bolts and not checking the unthreaded length of the shank due to the specifications bolts are typically made to in SA. Just ordering grade 8.8 bolts and giving a length did not necessarily give you the right unthreaded shank.

THE RESULTS OF THIS PRACTICE
Well, we had no structural failures directly as a result of the bolt qualities supplied and installed (unless as did occur in a few cases Grade 4.8 bolts were used in error in connections where 8.8 bolts were called up.)

We did have some bolt failures usually discovered during the tightening of the bolts. The most common failure situation was when Grade 10.9S (HSFG) bolts in a hot dipped galvanized finish snapped due either to Hydrogen embrittlement or being much too ‘hard’ so that they were brittle, with occasionally bolts being so ‘soft’ that they stretched too much in the tightening process. But in general we managed quite well.

What is the impact on our industry in general when structures are designed to European design codes and norms?

One of the first major differences we note when looking into the design and fabrication to the EN suite of documents is that the technical and quality requirements for steel structures designed to EN go up as the ‘execution class’ goes up.

There are 4 classes – EXC1, EXC2, EXC3 and 4.

To determine the class applicable one would consider the ‘consequence classes’ (risk profile to humans, economic and environmental failure), hazard profile based on things like stress levels in members, service factors (service categories), complexity of fabrication factors (production categories taking account that higher strength steels are more difficult to work with (460 Mpa yield to 900 Mpa yield)) etc. and by plugging the right numbers into a matrix out pops the Execution Class (EXC).

 EXC1 relates to the simplest of static steel structures, in low seismic areas, using steel grades lower than S355 and would have largely low specification welding requirements. (In SA this would be your typical security type works, car ports and the like)
 EXC2 is the default standard, most of our typical day to day structures would fall into this category.
 EXC3 would be used for dynamic structures with high consequences of failure, bridges etc. subject to vibrations, using steel Grade S355 and higher with important welds done on site including circular hollow section work with developed end welded connections.
 EXC4 would apply to only structures that would have extreme consequences and in Europe would be defined by legislation.

WHAT DO THESE EXECUTION CLASSES MEAN TO THE FABRICATOR

I will be so bold as to state that for EXC2 and EXC1 there is very little difference between the EN requirements for quality and what our typical SA workshop with an effective quality management system is currently doing using quality plans.

But when we step up to the requirements for EXC3 (Including the power stations!) and EXC4 then the rules become quite demanding…
 For example for EXC3 and 4 standards all constituent materials will be traceable back to source (even the bolts!) This is something that only the most specialised of work called for in SA!
 Or for EXC2 holes may be punched full size through the steel subject to the diameter being greater than the thickness but for EXC3 & 4 punching full size is not permitted i.e. holes are to be punched 2mm undersize and reamed out to the correct diameter.
 For all classes there is a hole tolerance of 0.5mm on diameter!
 For EXC3 and higher structures all bolts are to be ‘pre-loaded (pretensioned)’ bolts.
 For welding requirements, EN3834 management system registration is a requirement, i.e. to part 3 for EXC2 and to part 2 (the comprehensive requirements) for EXC3 & 4.
 The acceptance criteria for defects in welds are dependent on the execution class. As the class goes higher, the fewer defects are allowed.

LET US NOW CONCENTRATE ON THE BOLT REQUIREMENTS

The mechanical properties of all bolts to European specifications are now to be in accordance with ISO EN 898:1 of 2009. “The mechanical properties of fasteners made from carbon steel-bolts screws and studs with specified property classes – Coarse thread and fine pitch thread”. The property classes are from grade 4.8, 8.8 10.9 and 12.9 (there are others but we do not get them in SA).

The 1988 version has been used in South Africa for quite some time. It was also published under the guise of SABS1700-5-1:1996. There was a list of 10 possible tests described in the document. The hardness requirements are specified without any emphasis on their importance (hence not listing them in the Red Book). The 2009 version has very strict requirements for the mechanical properties of the bolts especially the hardness requirements. The list of possible tests has grown to 15! It covers the full range of diameters. It does not cover dimensional requirements. The specification clearly calls up all the mechanical properties of the material including ultimate tensiles, elongation, hardness, Charpy and the like.

The specification then defines which of the 15 possible tests can be carried out on the material (i.e. the finished bolts or parts thereof), what the test is for and acceptance criteria.

The following ‘definitive extract’ relates to the hardness issues: Par. 9.9 contains the hardness test requirement. It starts off with the following: “the purpose of the hardness test is for all fasteners that cannot be tensile tested-to determine the hardness of the fastener (which shall fall within the range specified in table 3 i.e. this is used to determine the suitability of the bolt or otherwise.” So in addition to providing for a test suitable for short length bolts not previously covered, it provides a way of testing bolts when you cannot do tensiles, charpy, elongation and the like such as when a bolt that is already installed and tightened is suspect. This has resulted in a great emphasis on hardness tests when there is reason to suspect the quality of the bolts.

Quote from the spec “- for fasteners that can be tensile tested – to determine the hardness of the fastener in order to check that the hardness is not exceeded.”

Hardness can be determined on a suitable surface (prepared suitably) or on a transverse section 1 diameter in from the end of the bolt. In a dispute relating to hardness readings only the latter readings are definitive.

Please go to Part 2 of this article for more technical details of the requirements of ISOEN 898:1

I believe we have covered most of the important issues / differences for the mechanical properties for the material for the different class grades as required for the bolts, what are the requirements for the bolts themselves?

ISO EN 14399 PARTS 1 TO 9:2005 – HIGH STRENGTH BOLTING ASSEMBLIES FOR PRE-LOADING
This is the specification that covers the bolts themselves (i.e. dimensional requirements etc.), calling up ISO EN 898 for material requirements and especially what makes for a usable assembly (i.e. bolt, nut and washer(s)) for preloaded bolts (either in shear or friction grip).
 Part 1 covers general requirements
 Part 2 covers suitability tests for pre-loading
 Part 3 covers system HR (described below) bolt and nut assemblies
 Part 4 covers system HV (described below) bolt and nut assemblies
 Part 5 covers plain washers
 Part 6 covers plain chamfered washers
 Part 7 covers HR system countersunk head bolt and nut assemblies
 Part 8 covers HV system hexagon fit bolt and nut assemblies (close tolerance in SA)
 Part 9 covers direct tension indicators (load indicating washers in SA)
 Part 10 covers TC bolts (torque control bolts in SA)

THE GENERAL REQUIREMENTS AND INFORMATION ABOUT PART 1

The introduction and Table 1 put the whole 9 part specification into context.

“This document on structural bolting reflects the situation in Europe where two technical solutions exist to achieve the necessary ductility of bolt/nut/washer assemblies. These solutions utilise different solutions (HR and HV described below) of bolt/nut/washer assemblies as described in Table 1 (below). Both systems are well proved and it is up to the experts responsible for structural bolting whether they use one or the other system.”

NOTES REGARDING THE ABOVE INTRODUCTION TO 14399-1:
1. Pre-loading means the same as pre-tensioning – the term we often use in SA.
2. HR system is the (new) European name for what we in SA used to call HSFG bolts with the name 10.9S. the old specification has been withdrawn and is replaced by EN14399).
3. HV system is a shear/ bearing type connection where pre-loaded bolts (pretensioned bolts) are used.
4. The decision of the expert referred to above comes out of the assessment done using Euro design codes leading the expert as to whether or not pre-loaded connections for the whole structure will be used or not. In South Africa we would have decided by joint type as to whether we need pre-loaded bolts i.e. (There are very specific applications where a structural engineer would use preloaded structural connections. He is the expert who must decide as inferred by the foreword to 14399)      – For a bolt with an expected tension in the bolt we would suggest an HV system


– In a vibrating structure HR or HV would be used
– For a non-slip connection we would suggest an HR system (HSFG)
5. It is important to note that for shear/bearing connections designed to SANS 10162 in SA we always assume the shear plane will be in the threaded portion of the bolt (i.e. the weakest part of the bolt). In Europe the shear plane is always designed to be in the unthreaded shank of the bolt. Threads in the shear plane are not allowed in accordance with EN14399 for structural design for pre-loaded applications. This is in conflict with what has been standard practice for designs done in SA.
6. In both parts 14399 3 and 4 which cover the actual dimensions of the bolts and nuts, note that the head of the bolt is ‘thickened up’ under the head for most of the bolt head area (see Diagram 2 on page 38). In terms of clause
4.4.2 the under head radius shown in the detail X per figure 1 of 14399 3 and 4 is required in class 10.9 bolts to reduce susceptibility to hydrogen embrittlement. The Europeans have also engineered out the possibility of hydrogen embrittlement by adopting Zinc Thermal diffusion coatings and eliminating the use of acid for any cleaning operations. Again this places the requirement for the hardness/tensile test as being very high as well as material selection which has been inadequately dealt with previously.
7. Washers to 14399 part 5 are flat washers and intended to be used under the nut where required. The washers to 14399-6 have a chamfer to both the inner diameter and outer diameter on one side of the washer and intended for use under the head of the bolt to clear the radius at the joint of the head of the bolt and the shank.
8. Some of you may be familiar with clause 4.5.1.6. of SANS 2001-CS1:2005 “Galvanized nuts Nuts that are to be hot-dipped galvanized shall be of a higher class than the associated bolt or screw”

Please be advised that there is no such requirement in 14399 and in future updates of SANS2001–CS1
– this requirement will be dropped.

The issue of pre-lubricated nuts, which is a big issue and part of the whole reason EN14399 was introduced has not been covered by this article but will form part of a further article relating to tightening issues for bolts.