Classical ancient Greek architecture, as exemplified in temples such as the Parthenon, is characterised by the use of pillars and lintels. Stone is of course strong in compression (so good for pillars) but weaker under tension and bending (so not good for long unsupported lintels). Simple stone beams or lintels cannot safely be used to span distances of more than about 3 metres otherwise they are liable to crack. Therefore in order to provide practicable roofs for Greek temples and other buildings it was necessary to use wooden beams, but even wooden beams will sag if not supported. Hence the inside of a large temple would have been crammed with a forest of supporting stone pillars. As I understand it these roof beams supported a ceiling of planks on which was piled packed earth to make a waterproof covering, or later light timber slats supporting tiles. But all these roof coverings were basically lying one on top of the other, relying simply on their weight to stay in place on the supporting pillars and lintels, and so were in no way structural.

As JE Gordon in “Structures or Why Things Don’t Fall Down”, says:
“The brilliance of Hellenic architecture seems to come to a stop, rather suddenly, when one gets to the architrave. Greek roofs can only be described as intellectually squalid.”

Arches were to be the way forward, and it is often pointed out that the Greeks seem not to have known about arches and so it was left to Romans to build the first wide unsupported interior spaces, such as in the Pantheon in Rome. Strictly speaking the Greeks did know about corbelled arches. Such arches occur in Mycenae and the vaulting of the passages within the walls of Tiryns (all built well before the time of Homer) but such applications were only used in situations where the side thrust could be supported, either by massive masonry or by earthworks. And anyway even when the sideways thrust is adequately supported a corbelled arch has a very limited horizontal maximum span in comparison to its vertical height.

To the modern mind it might seem fairly clear that the most promising way to try to bridge a roof-span using short pieces of timber would be to join the short members together, Meccano-fashion, so as to make a triangulated lattice structure. Such a truss behaves like a long solid beam, but is actually more resistant to sagging than a simple beam. If properly designed and constructed even a very simple lattice truss can span a considerable roof span, and moreover it does so without putting any, or very little, dangerous outward thrust upon the walls, which is the eternal problem with masonry arches. Roof trusses might seem obvious but they appear to have been surprisingly slow to be developed. Architectural trussing seems to have been a late Roman invention and only really caught on in the Middle Ages.

But the Greeks DID know about trusses:

Ancient Egyptian boats, being constructed from papyrus, were very flexible and so to stop the bow and stern bowing outwards ie hogging, they were trussed by means of a tensioned rope joining prow to stern: a hogging truss. Greek ships were considerably more advanced than this, but due to the scarcity of timber were built from relatively short planks, and consequently their light hulls could be alarmingly flexible. For this reason the Greeks retained the hogging-truss in the sophisticated form called the hypozomata: a substantial rope about 50mm in diameter and twice the length of the ship. In the modern reconstruction of a trireme, the Olympias this rope runs from the bow down the centre line of the vessel under the decking to the stern-post, and then back to the bow. But I have also seen it suggested that originally the rope would actually run right around the outside of the hull, high up just beneath the gunwale.  Whatever its exact position it is clear that the hypozomata, whose tension could be adjusted as needed by a windlass, functioned as a sophisticated truss and served to turn a flexible hull into a much more rigid structure, capable of being rammed at speed into an enemy ship like a missile.  

The Greeks, not generally slouches when it came to mathematics, engineering and innovative thinking, clearly understood the concept of the tensioned truss. But for some reason they didn’t exploit it in building where it need not have been anything as sophisticated as an adjustable rope but just as simple, at least as a starting point, as three planks securely nailed together at their ends to form a sturdy triangle. But why didn’t Athenian architects latch onto the same idea for the roofs of their temples? If they had adopted the ship-stiffening technology in architectural building to create A-frame roof trusses, they could have greatly increased their possible roof-spans without actually using any more wood, and in particular without needing to use any scarce longer lengths. Nor as many eye-wateringly expensive carved marble columns.

In spite of the limitations of stone beams, the size of ancient buildings was often greater than that of corresponding modern ones. The Parthenon, for instance, is considerably bigger than many churches and cathedrals. Nevertheless the Parthenon, about 70 by 30 metres, is small compared with Hadrian’s temple of the Olympian Zeus close by, which measures 108 by 52 metres and would fill most of Trafalgar Square.  They didn't generally lack artistic vision and they don't seem to have lacked the money, resources, nor logistics to regularly undertake and complete massive building projects. So just think what they could have done if they’d developed the roof truss!

But is there, perhaps, some evidence that they did use the roof trusses, if only for domestic buildings? And if they really did not develop even the simplest A-truss, are there any particular reasons why not? Was it just that they’d never played with Meccano and so couldn’t see the analogy between ships' keels hogging and roof-beams sagging?

Or was it simply that Athenian architects, such as Mnesicles and Ictinus, never hobnobbed with shipwrights?

Romans nothing - we Sumerians invented ("discovered" might be better) the arch, and our neighbours like the Babylonians copied us - this is the reconstruction of the Ishtar Gate.

Yeah, but that's just a corbelled arch hidden behind some fancy tiling, and so the actual span isn't much bigger, if at all, than what one could do with a simple stone lintel. It looks pretty but it isn't really wide-ceiling technology.

But one could then ask: if the Sumerians, Assyrians, Babylonians had really invented the true arch ... then why didn't they ever do really wide naves and huge domes, like, say, Hagia Sophia? I'm sure it wasn't for any lack of man-power nor a want of meglamania!   

Harriet Crawford would appear to be less convinced than you are. In "Sumer and the Sumerians" she suggests that the Sumerians "probably invented the true arch as well as the corbel vault and the dome" Mud brick might not be the best of materials, though, for large unsupported structures.

But that is beside the point. I completely accept that the true arch was invented/discovered in the Near East long before classical Greece ... But my original point was: given that they seemed to understand trussed structures (in ship construction), then why didn't the Greeks make the intellectual small step to use the same ideas to solve their contemporary problems in architectural construction?

And it is an interesting thought that, if the Greeks HAD indeed developed trussed and lattice structures - and there is absolutely no practical reason why they couldn't have built such things with the technologies available to them at the time .... then we might have actually by-passed arches and vaults, for all their ribbed, groined and barrelled, gothic majesty ... and gone straight to modern-style, wide un-supported lattice roof ceilings. Or maybe even to tensioned 3D tent-like structures like the Millennium Dome.

It's a good question. If you compare the Greeks and Romans then it is obvious that the Romans, for all their well-earned reputation for being the Japanese of their day (great imitators and developers of technology, maybe lacking a bit on the pure innovation front), still blithely transferred what they learnt in one field of engineering over to others as a matter of course, unlike their Greek counterparts. The canopied awnings over their arenas are much vaunted examples, owing everything to marine technology as they did, but surely knowledge of stress dissipation and gaining maximum constructional integrity with minimum mass of material was also a feature of this transition of knowledge, as useful to implement when building aqueducts as hulls. I reckon also that it was the aqueduct - monumentally sized structures required through necessity as opposed to decorative vainglory - which allowed the Romans to experiment with and refine arch technology at an expense so huge that it might have precluded them earlier, and the Greeks before them, from daring to employ such a trial-and-error perfection of the technique in the erection of public buildings, the only other area where it could feasibly have been done on such a scale.

Regarding the canopied awnings over Roman arenas and the cross-over between different environments/technologies that these represent ... that is a very good point about the Romans, who indeed get loads of stick for supposedly never inventing anything original themselves, but they do indeed seem to have been quite good at lateral thinking and developing the ideas of 'others'.

Whilst they, like the Greeks, were always quite keen to built for posterity in stone ... whether as pillars-and-lintels, or masonry arches/domes .... they, unlike the Greeks, do seem to have been more open to new ideas, such as trussed structures, if only for temporary or lesser domestic structures. Trussed wooden constructions certainly seem to have been employed, almost casually as if such things were already well-known, in the construction of wooden forts in Gaul and Germania. And further, though I have to rely on integrity of archaeologists here, many reconstructions of Romano-British villas show A-shaped, H-shaped, double V-shaped etc.. trussed roofs. Though I'm not sure whether these interpretations are based upon actual evidence or just reflect a modern assumption that Romano-British roofs were (obviously!) constructed just like medieval British roofs? Although I doubt anyone has yet dared produce a reconstruction of a late Roman villa with a characteristically-English, medieval-style, hammer-beam trussed roof!

7 May 558, ( a Tuesday) the dome of the Hagia Sophia in Constantinople collapses;

http://www.learn.columbia.edu/ma/htm/or/ma_or_gloss_collapse.htm

It is noticeable that in the two catastrophic failures of the Hagia Sophia dome mentioned one was the result of the incineration of wooden components during an assault by rioters (rectified in the rebuild by using only non-combustible materials) and the second was stupidity on the part of renovators who removed mass from the structural supports dissipating the stress of the dome. In both cases the dome's structural integrity before sudden artificial change had not proved lacking in the slightest - in the second it had even demonstrably survived the great earthquake of 557 which Agathias reported had "almost completely razed Constantinople to the ground".

MM,

while looking to reappearance of bricks in Nothern Europe in the 12th century I found your post here.
I had to look for "trusses". As I could guess from your description I had already some ideas...
And I found for "trusses" in Dutch: spanten, gebinte and so I had a better insight with what you meant.
But it seems to be also the Dutch "trossen" as trusses bananas or grapes  
https://en.wikipedia.org/wiki/Timber_roof_truss

Kind regards from Paul.

Blimey Paul, I'd completely forgotten I'd started that thread and clearly it was written from the same notes I used for my initial post about Greek arches. I'm glad you found it as I was about to post something very similar about roof trusses on the 'arches' thread.    What's it called when you keep forgetting things?

Actually the Greeks need not have had to have experience of anything as sophisticated as the hypozomata - the tensioning rope they used to stiffen their ships - as a simple archer's bow should also have indicated the concept - although nevertheless they did still fail pick up on its significance. Of course tensioning a roof structure with rope was not then really very practical (unless they they'd been able to produce steel wire), but a few nails to join all the roof timbers together into a sturdy A-frame was certainly not beyond their capabilities. However they didn't even take that step and continued to build much like with a child's building blocks, by simply relying on the weight of the stack to stop it all from falling apart.

I also see that at one point in the above I dismissed the Ishtar Gate as 'just' a corbelled arch faced with fancy tiles - but it is actually a true arch entirely constructed from kiln-fired bricks.

Meles meles wrote:

Trussed wooden constructions certainly seem to have been employed, almost casually as if such things were already well-known, in the construction of wooden forts in Gaul and Germania. And further, though I have to rely on integrity of archaeologists here, many reconstructions of Romano-British villas show A-shaped, H-shaped, double V-shaped etc.. trussed roofs. Though I'm not sure whether these interpretations are based upon actual evidence or just reflect a modern assumption that Romano-British roofs were (obviously!) constructed just like medieval British roofs?

I was browsing the National Trust website earlier and noticed on their Picnic Spots section that Mussenden Temple is open today only to tickets holders for the Downhill Gun Salute. I don’t know what the Downhill Gun Salute is, and the Events section offers no information as to what exactly is being saluted on the 20th of August. I’d guess that it’s got something to do with either the history of the ruined Downhill Demesne estate or else the life of its 18th Century founder Frederick Hervey, Earl of Bristol.
 
Not that today I intended to make a day trip to County Londonderry, but I have been to Mussenden Temple before and its setting on the cliffs above Benone beach overlooking the Atlantic is truly spectacular. It’s also something of a marmite feature in the landscape. Some see it as an inappropriate intrusion and the epitome of the aristocratic follies built in that era, while others see it as enhancing same. Even in its day, there were those who criticised its choice of location and architectural style even correctly predicting that the saline winds would one day destroy it. In this they were correct, although the original Downhill House was actually destroyed by fire before being abandoned in the 1940s. Ironically, the most exposed building on the estate, the Temple, survives. And it wasn’t technically a folly at all but served as Hervey’s library complete with its own hypocaust to protect the building and its books even on the darkest and wettest of winter nites.

This got me to thinking, however, that if the ‘Edifying’ Earl-Bishop Hervey was able to commission the building of a Roman-style dome supported by a rotunda with decorative Corinthian columns in the 1780s and using local labour in Ulster, then did the Romans themselves ever build domed buildings in Britannia. I can’t think of any examples. Indeed, there are very few examples of domes having been constructed anywhere in the northern part of the former Roman empire with exceptions perhaps at Augusta Treverorum (Trier) and Colonia Agrippina (Cologne).  



(Mussenden Temple teetering on the cliff edge. Note the sash windows still glazed and intact.)

I visited Northern Ireland four times in five years 1999-2003 and took in such wonderful places as the Mussenden Temple, the Carrick-a-Rede rope bridge, the Giant's Causeway and my favourite - Dunluce Castle. I also stayed in Belfast, Newcastle and Enniskillen.