Visual grading of Existing Structures

For a much more in-depth view on the topic: Mike Bather’s thesis on The assessment of the mechanical and physical properties of in situ timber.

National and International Standards provide useful regulations to ensure the efficient trade of goods, customer safety and quality control. However, in timber grading some standards are used for purposes that they were not intended to cover. In the assessment of existing structures standards for the visual grading of timber are sometimes used as a strength assessment method. Is this a good idea?

Let’s take the British Standard BS 4978 as an example. It lists a number of visual grading criteria that should help to assess load bearing properties of softwood timber. With the help of these criteria, timber can be sorted into two categories, “General Structural” and “Special Structural”, or can be rejected for use in construction (if you are not familiar with the standard, find it explained here). The standard was designed for virgin timber and some problems arise when the strength of existing structures or recovered timber should be assessed. Nonetheless the standard is sometimes used to estimate the strength of timber members in historic buildings. In lack of any other British standard for this situation, using the BS 4978 rules might seem like a reasonable approach, but not only was this standard never meant for grading non-virgin timber, but grading rules are never meant for evaluating individual timber members either (If this seems strange to you, read more about grading in general here). But should these problems stop people from using the standard for recovered timber and assessment of existing structures? And are there better solutions available?

Visual assessment

The visual grading criteria like knots and ring width are only part of the evaluation of timber. The standard only allows the assignment of a strength class when the species and growth region of the timber are known. In existing structures it is very unlikely that information on this is available. Even if it was, the properties of timber used in historic buildings are very likely to differ from the timber used as a reference. That means we should probably not use the standard to assign a strength class. But this is not dramatic, since strength classes are only needed to put timber on the open market and are not, as many people seem to forget sometimes, characterising a piece of timber.

Obviously, in an existing structure you “only” need to proof that an individual timber member is fit for its purpose, a strength class assignment is not needed. For evaluating the strength of a piece of timber the visual grading criteria from the standard might help, but rules and common sense should not be used mutually exclusively. Next to individual characteristics of the specific case, some particularities in existing structures should always be taken into consideration:

1. Not all criteria might be used due to limited visibility and accessibility
2. It might be useful to focus on the most critical points for some of the criteria, e.g. wall ends for fungal decay and high load zones for cracks
3. Fissures are likely to occur more frequently and tend to be bigger due to drying effects (meaning the standard criteria will seldom be met), but drying fissures are not usually a problem
4. Wane and geometric deviation from the rectangular cross section might be more severe in historic buildings but are not necessarily strength limiting factors
5. The actual use case of timber members is known, so the load scenario can be taken into account when evaluating degrading features
6. Interaction of different members in the structure influences the experienced loads and so the relevance of defects

In case you can get an unobstructed view of a timber member, at least some of the BS 4978 criteria might be of help in the assessment of strength. The growth rate is the only visual criterion that helps to estimate density, but the correlation is usually not very strong. If additional non-destructive tests are used for density measurement, growth ring width is unlikely to yield useful information. Knot size is not a great predictor for strength either, but still people use them to assess the “grade” of the timber. Decay is likely to diminish strength and its extend should be evaluated, but the grading standard does not give any indication on the quantification of the strength decrease, it just tells you to reject pieces with decay. But we can’t really reject timbers that are already in a structure. Especially in important historic structures it might not be a sensible approach to replace every member that fails a visual grading (to rules that were never meant to assess old timber!). So the extent of decay, fissures, wane etc. should probably not be limited quite so strictly in historic structures. At the moment the assessment of such defects often comes down to the gut feeling of the expert inspector.

The visual grading rules can help with strength estimations, but in addition it is important to evaluate damage that is not covered in the British standard. If fissures, wane and decay are present, the remaining cross section should be estimated and updated for the calculations. Some national standards are focused on defects in existing structures and can be used as guidance: SIA 269/5 (Switzerland), UNI 11119 (Italy) and, to some extent, the international standard ISO 13822.

Still, none of these standards provides an easy-to-use-recipe for strength assessment. An expert judgement of the individual case is always needed. If this includes some grading criteria from BS 4978 (or one of the above mentioned standards), that’s fine, as long as the limitations of the standard are kept in mind.

Using additional tools

Different methods can be used to increase the confidence of a visual assessment: resistance drilling, core drilling, ultrasonic measurements, X-ray, thermal imaging and more non- and semi-destructive methods are being used for different assessments. Of course, interpreting the results of these tests requires some training and experience. And the tools are not always cheap or easy to handle. It is also very hard to estimate strength from any of these methods, while density and stiffness can be predicted more reliably. These limitations of the assessment tools provoke that most assessments will actually be done with visual and simple mechanical methods (à la “If I can stick a screwdriver halfway into a beam, it’s probably not safe”).

A non-destructive, quick and possibly important test is for the moisture content of wood. A handheld moisture meter can be useful to detect dampness where fungi and insects can attack, since even wood that feels dry to touch can be at risk. If the moisture content of timber is above 20%, measures for moisture protection should be considered.