…..This comparison with medicine fits very well, there are really many similarities between the work of the technician who has to understand what to do with an existing building and that of the doctor who tries to make a diagnosis and to find a correct therapy for a patient…..
versione italiana qui: Gli edifici in muratura alla prova dei terremoti italiani (colloquio con Guido Magenes)
Guido Magenes is Professor of Structural Engineering at the University of Pavia and IUSS Pavia. He is also the coordinator of the Masonry Structures division of the EUCENTRE Foundation. His area of greatest competence is the seismic behavior of masonry buildings and for this reason he has also participated and still participates in numerous Italian and European technical-regulatory committees.
We discussed with him the behavior of masonry buildings in Italy, with particular reference to what happened during the last earthquakes.
1. The earthquakes of 2016 have determined a sequence of shaking that has put a strain on the buildings of the affected area, especially those in masonry. The effects seen in the field are very different: next to the buildings already heavily damaged by the earthquake of August 24th, there are others that have seen their condition worsen after the shock in October, and others that seem not to have suffered serious damage in all the sequence. Do you have an explanation for this?
The masonry buildings stock in our country has very variable characteristics and qualities, depending on the era of construction, the materials and construction criteria that were used, the type and architectural form (ordinary buildings or churches, palaces, towers, etc … ), any maintenance and reinforcement or tampering and weakening processes that may have occurred over time. Certainly there are recurrent types of problems, but the diversity of the behavior of masonry buildings, apart from the severity of the shaking (or the different ground motion in the various sites), is essentially due to this great variability.
Therefore, in the specific case of the seismic sequence of central Italy, which involved a very large area and a considerable variety of buildings, we observed what you say: from the recently built building, of a few storeys, in great part or fully compliant with the modern design and construction criteria, which did not show significant damage, to historic buildings with large spans and heights, such as churches, which tend to be more vulnerable and have therefore suffered great damage and collapse because of their dimensions, geometric ratios and their structural organization. In many if not most cases, also the poor quality of the materials has further worsened the situation.
2. In all the municipalities affected, seismic regulations were in force, with various years of enforcement (the extremes are represented by Amatrice and Accumoli, 1927, and Arquata del Tronto, 1984). The distribution of the damage does not seem to be influenced by these differences; is there a reason?Not all regulations are equally effective: a 1927 standard is obviously very different, under many points of view, from a rule of the 1980s or the years 2000s and, as I mentioned above, the buildings built in compliance with the latest rules behaved generally well (constructed with artificial blocks and mortars of good strength, or even stone buildings demolished and rebuilt with good quality mortars). Therefore, I would not say that the distribution of damage is not at all influenced by the regulatory context. It depends on what was written in the norm and how many buildings were built or repaired or reinforced after the introduction of the norm (in the affected centers a significant percentage of the buildings had been built before the seismic regulations that you mentioned).
The rules and design criteria are not necessarily born perfect and they have to adjust, to evolve based on the experience of earthquakes. For example, it is only fifteen or twenty years that we began to recognize that certain types of interventions proposed and widely applied after the earthquakes of Friuli and Basilicata can be harmful or plainly ineffective (think of the infamous reinforced concrete ring beams “in breccia” inserted at intermediate floors in an existing building in stone masonry: in Umbria-Marche ’97 we have begun to see its shortcomings).
In the areas in which the presence of a regulation or a seismic classification seems to have had no effect, it must also be taken into account that the on-site control of the quality of construction and execution, in particular for masonry buildings, were inexistent or ineffective at least until the more recent regulations. The use of a very bad mortar is a recurrent element in many of the old masonry buildings collapsed or damaged in the last seismic sequence. In centers like Accumoli and Amatrice it seems that even where interventions had been carried out on buildings, replacing old floors, for example, or inserting some ties, the problem of poor quality of the masonry had been greatly overlooked, ultimately making the interventions ineffective. We can add that a large part of those areas suffered a considerable depopulation since the early 1900s, with inevitable consequences on the maintenance of buildings, which has led to an increase in widespread vulnerability.
Then there are some particular cases in which historical norms and more recent norms seem to have had a positive effect. Take Norcia’s example: without going into the details of the measurements of the characteristics of the ground motion, it is a fact that Norcia in the last sequence suffered strong shaking, comparable to those of Amatrice and Accumoli but with a much lower damage to buildings. In the history of Norcia there were two very significant events that may have affected the response of the buildings in the 2016 sequence, one before and one following the 1962 regulations. In 1859 a strong earthquake caused numerous collapses and victims in some areas of the historical center, and following this the Papal State issued a quite effective regulation that gave a series of provisions for repairs and reconstructions: on geometry, in particular on the maximum height of the buildings (two floors), on the construction details, on the quality of materials. Then, in 1979 there was another earthquake in Valnerina, after which other parts of the historic center were damaged, followed by a series of systematic reinforcement measures on many buildings. In many of these buildings the reinforcement of the vertical walls (even with the controversial technique of the reinforced plaster) has remedied one of the main elements of vulnerability, i.e. the weakness/poor quality of the masonry walls. If for a moment we leave aside the elements that can go against the use of reinforced plaster (such as the durability of the intervention), and we see it simply as a technique that has remedied a factor of great vulnerability, we can say that for Norcia there has been a positive combined effect of pre-modern and more recent regional regulations, stemming from the direct experience of seismic events.
3. Let’s talk about seismic regulations and in particular of their engineering aspects. We hear that they have changed a lot over time, and that perhaps the non-recent ones were not entirely effective. Is it true, and if so why?
As for the engineering component of the regulation, what we now know about the structural and seismic behavior of buildings, in masonry and other structural systems, is the result of a continuous evolution through the experience of earthquakes in Italy and in other parts of the world. In Italy the engineering study of masonry buildings has resumed life, after decades of almost total abandonment, after the 1976 earthquake in Friuli. The first norms/codes that give indications on how to “calculate” a masonry building in Italy date back to the early 80s (to “calculate” I mean “quantitatively assess the level of safety”). Although “calculation” is not the only component of the design, this fact gives the idea of how only the very recent rules have a technical-scientific basis aligned with current knowledge. I would like to say that the absence of calculation in a project does not necessarily imply that the building is unsafe: in the past we followed geometric and constructive rules of an empirical type, based on the experience and intuition of the mechanical behavior, although not explicated in detailed calculations. Even today, for the design of a simple and regular masonry building, it is possible to follow codified geometrical and constructive rules that avoid detailed or complex calculations, but still achieve an adequate level of safety.
The experience of the earthquakes of Irpinia, Umbria-Marche, until the most recent in central Italy, have been a continuous test and a source of knowledge. For example, as mentioned in my answer to the previous question, the Umbria-Marche 1997 earthquake, besides highlighting the great vulnerability of churches and of certain historical structures, has been an important test for strengthening criteria and techniques on masonry buildings that were proposed and developed following the Italian earthquakes of the late ’70s, showing how some techniques are not very effective or can even be harmful if applied indiscriminately and without awareness
To conclude my answer with my opinion on current technical standards, I think that as regards the design of new buildings we are really at a very advanced state of progress, which effectively attains the levels of safety that today are considered adequate. I think there are more uncertainties on the assessment and strengthening of existing buildings, even if it is not so much a regulatory problem but rather of scientific knowledge and of the correct identification of strategies and techniques for the intervention. It is certainly easier to design and build a seismic-resistant building from scratch, than to assess and intervene on an existing building.
4. How much – and how – does the construction and detailing of a building affect its seismic safety, beyond the design?
The question gives me the opportunity to dwell a little more on what is meant by “design”, which is something different from the mere “calculation”. The design includes all aspects of overall conception, choice and organization of the structure, choice of materials and construction techniques (with the awareness of how they can and should be executed in situ), performance verification calculations in terms of safety against collapse and of satisfactory behavior in normal operation, prescriptions on construction details. In modern seismic design it is also necessary to take into account, when relevant, the seismic response of the non-structural parts of the construction. There must also be a check that what is prescribed in the design is actually implemented during construction.
The calculation is therefore only a component of the design. It is interesting to note that most of the existing masonry buildings were not calculated, at least as we understand structural calculations now. The first Italian national technical standard on masonry constructions with a sufficiently detailed description of the calculations for the structural verification dates back to 1987. Technical standards with indications for the seismic calculation, were issued after the earthquake of Friuli 1976 and in subsequent times. Before those norms, a technical literature and manuals existed, with reference to the principles of mechanics, as well as a building tradition. I would like to clarify that here I am talking about regulations/norms that tell how to calculate the resistance of a masonry building, subject to seismic or non-seismic actions. Just to give an example, the Royal Decree of 1909 (post earthquake of Messina), a historical milestone as regards seismic regulations, gives criteria to define the seismic action, gives constructive and geometric rules but does not tell how to calculate the resistance (the capacity, according to the modern technical language) of a masonry building.
The constructive tradition based on the respect of the “rule of art” always had in mind the importance of construction details, of the quality of the materials, of how the building is built, and this has allowed and allows well-constructed buildings (but not “calculated”, i.e. non-engineered) to withstand even very violent seismic shocks. In modern buildings, the compliance during construction site of the execution rules, the control of the quality of the materials, is equally important, although this holds for masonry as for the other types of construction. The sensitivity of the structure to constructional defects is a function of the level of robustness of the structural system. A masonry box-like construction, strongly hyperstatic (i.e. where the number of resistant elements is higher than the minimum necessary to ensure equilibrium under the applied loads) could in principle be less sensitive to construction defects than an isostatic prefabricated structure (i.e. where the number of resistant elements is just equal to the minimum necessary to ensure equilibrium under the applied loads, so that the failure of a single element is sufficient to generate a collapse). Obviously we are talking about local defects and not generalized over the whole construction. If all materials are poor quality throughout the construction, then it is a great problem, but not necessarily a masonry building is more sensitive to such problem than, say, a reinforced concrete frame, in which also defects in the reinforcement detailing are possible (for instance in beam-column joints or in lap splices or in anchorage of rebars and so forth).
5. Many surveyors in post-earthquake reconnaissance activities have found traces of interventions that have allegedly weakened the structures. Do you agree?
In post-earthquake surveys, carried out quickly in emergency conditions, it is not always possible to clearly understand the history of the building and what changes have been made, in what time and modalities, but sometimes it is clear that some modifications to the construction have been detrimental to safety. Often these are interventions that were made with total unawareness of the effects on structural safety and purely for the purpose of use and redistribution of space. In other cases, more rare, there are also interventions made with “structural” purposes, perhaps even with the idea of achieving an increase in safety, but which in reality were harmful or ineffective. A classic example, often discussed in the literature also on the basis of the Italian post-earthquake recognitions from Umbria-Marche 1997 onwards, is the insertion of new, rigid and heavy structural elements (such as the replacement of a wooden floor with a reinforced concrete floor) in a building with very weak masonry (for example masonry made of irregular stones with poor mortar), without the masonry being properly consolidated. There was a period, following the earthquakes of Friuli and Irpinia, where much emphasis was given to the fact that rigid diaphragms (i.e. the floors and roofs) increase the hyperstaticity, hence the robustness of the construction and the so-called “box behaviour”, by which engineers tried to replicate in existing structures something that is relatively simple to implement, and whose effects are well controllable, in new constructions, but which in an existing construction has great problems of practical implementation (particularly in the connection between new elements and existing elements) and of potentially negative structural effects (increase of stresses in an already weak masonry). It is important to note that the effectiveness of the interventions is tested by earthquakes that take place in later times, and in some areas of central Italy it has been possible to draw indications of this kind. In the earthquake of Umbria-Marche in 1997 it was possible to observe various problematic situations in buildings where the existing floors had been replaced by heavier and more rigid slabs.
Allow me, however, to add a further comment. From the scientific point of view, the fact that an intervention is “harmful” or weakens the structure compared to the non-intervention is verifiable experimentally only if there is a confirmation of what would happen to the building without intervening and what would happen following the intervention . This type of comparison in the vast majority of practical cases is not possible, except for very fortunate cases of almost identical buildings built on the same ground where one was reinforced and the other not, or that were reinforced with different methods. Or through laboratory experiments, comparing specimens tested on a “shaking table” (earthquake simulator). So, in general I am always rather skeptical of interpretations given on the basis of purely visual rapid surveys, without the necessary in-depth study of the details and without a quantitative analysis carried out in a competent and thorough manner.
I can say (and I know that many colleagues have a similar opinion) that in many cases seen in central Italy the collapse of the construction would have taken place regardless of the type of floor, light or heavy, rigid or flexible, by virtue of the bad quality of the masonry, which appeared to be the main problem.
6. How did the repetition of the strong shocks play in the aggravation of the damage (where it occurred)? Is it something that is implicitly foreseen, and taken care of, by the seismic norms? On the other hand, how do you explain the numerous cases of almost total absence of damage?
The repeated shaking aggravates the damage, the more the damage caused by the previous shock is serious. It seems a rather obvious statement, but essentially it is what happens. For example, if a first shock on a masonry building generates only a few cracks, not very wide and of a certain type (for example horizontal cracks, which close after the shock due to selfweight), the building has not lost much of its resistance; so if it is subjected to repeated shaking, less intense than the first shock, it is possible that the damage does not get too much worse, and if it is subjected to a shaking stronger than the first shock it will have a resistance equal to or slightly less than it would have if the first shock had not been there. On the other hand, if a shock leads to the development of diagonal cracks (so-called “shear cracks”) or vertical cracks with spalling, the damaged part has lost a significant portion of its ability to resist and subsequent repeated shaking can lead to progressive degradation and collapse, even if the subsequent shocks suffered by the building, individually, are perhaps less strong than the first one. This is something visible and reproducible also in the laboratory.
That said, there are types of constructions and structural elements that are more or less sensitive to the repetition of the seismic action. When seismic engineers speak of “ductility” of the structure or of a mechanism, they also refer to the ability of a structure to resist repeated loading cycles well beyond the threshold of the first crack or the first visible damage, without reaching collapse. A well-designed modern reinforced concrete construction is a structure of this type, for example. Unreinforced masonry, on the other hand, is more susceptible to damage induced by the repetition of loading cycles beyond cracking. As a consequence, existing masonry buildings, once damaged by a first shock, are more vulnerable to subsequent shocks. On the other hand, if the first shock does not cause significant damage, the safety of the building remains, in most cases, more or less unchanged and this accounts for the fact that numerous masonry constructions have also resisted repeated shocks. Unfortunately, sometimes the damage may not be clearly visible. Damage in masonry originates in the form of micro-cracks (not visible to the naked eye) which then develop into macro-cracks. If in a laboratory test a sample of masonry is pushed to a condition very close to the onset of the macro-cracks but the load is removed just before they develop, it may happen that in a subsequent loading phase the macro-cracks develop at a load level lower than that achieved in the first phase. It may therefore happen that a building that has resisted a violent shock without apparent damage is visibly damaged by a subsequent shock less violent than the first.
You ask me if the behavior of the structure under repeated shocks is implicitly considered in the seismic norms: the answer is yes, at least for certain aspects. For example, the respect of certain construction details in reinforced concrete and the application of certain rules in the sizing of the sections and of the reinforcement have this purpose: to make the structure less susceptible to damage under repeated actions. Moreover, less ductile structures, such as those in unreinforced masonry, are designed with higher seismic “loads” than the more ductile structures, also to compensate their greater susceptibility to degradation due to repeated action. However, there are some aspects of the problem of resistance and accumulation of damage under repeated shaking that remain to be explored and are still cutting-edge research topics. In particular, if it is true that theoretical models are becoming available to assess how the risk (i.e. the probability of collapse or damage) evolves in a building or a group of buildings as time passes and seismic shocks occur, these models must still be refined to give results that are quantitatively reliable.
7. It seems to me that the variety of masonry buildings, at least in Italy, is really large: so large that knowing them requires an approach similar to that of medicine, where each case has its own peculiarities. Therefore, there is perhaps no universal therapy, every case requires a specific care: is it correct? And if so, given that the building and construction techniques and quality of different areas of the Apennines (and others) are similar to those of the areas affected in 2016, should a similar destruction be expected to repeat again?
This comparison with medicine fits very well, there are really many similarities between the work of the technician who has to understand what to do with an existing building and that of the doctor who tries to make a diagnosis and to find a correct therapy for a patient. From the technical point of view there is no universal therapy and no (good) doctor would be able to apply a therapeutic protocol without the anamnesis, the objective examination, any necessary instrumental or laboratory tests and the formulation of a diagnosis (which tells us what is the patient’s disease / health status, and then defines what he needs, the therapy). The good technician follows a similar path to arrive at the evaluation of safety and possible hypotheses of intervention (or not intervention). Of course it is possible and necessary, as is the case for medicine and public health, to define strategies and policies for prioritization and allocation of resources to ensure that the overall seismic risk in our country decreases. Certainly, where the old buildings have not been subject to maintenance, or just to aesthetic and functional maintenance without structural reinforcement, we can expect destructions similar to those seen in 2016 on the occasion of future earthquakes of comparable magnitude. This applies to both public and private buildings.
Where instead we have intervened or will intervene in a conscious way, paying attention to the problem of seismic safety, the level of damage to be expected is lower, as the experience of the past earthquakes teaches us.
Allow me to conclude this interview with some non-purely technical engineering comments. The possibility of reducing the seismic risk in Italy depends on many factors, ranging from how politics govern the problem of natural hazards, to how technicians, individually and collectively, interact and communicate with politics, to how the presence of risk is communicated to the population, to how, as a consequence, the citizen makes his choices when he buys or takes decisions to maintain a property. In my opinion it is necessary to progressively evolve into a system in which the citizen recognizes that it is in his own interest to pursue a higher seismic safety, initially spending a little more, because he will have a return in the future not only in terms of safety but also of economic benefit, for example in the market value of his property. The “Sismabonus” initiative is certainly a first step in this direction, but other steps will have to be taken. The goal, certainly not easy to achieve, should be that the safety level of a building has a clear and recognized economic market value, and I think this would work for both the small owner and for real estate investors. I know that some are scared by this perspective, but personally I think that, at least for what concerns privately owned real estate and facilities, there are no other ways to achieve, within a few decades, a substantial and widespread reduction of seismic risk in Italy.