The siege was the most common form of battle in the Middle Ages. In most cases a siege consisted of little more than a larger army cornering a smaller force inside fortifications. The besieger could try and attack the fortifications, but this would often be risky (the assault would fail) and expensive (in terms of dead attackers). The most common tactic was to try and starve the besieged force out. This took time, and given the expense of maintaining a large besieging force over many weeks or months, the siege was often lifted because the besieging commander ran out of food to feed or money to pay his troops.
The besiegers faced a formidable task, as they had to take on places that were expertly built to keep people out. But a good reason to persevere was that there weren't that many fortified places. And the reason for that was the huge expense of building these places.
Even a simple stone tower could be incredibly expensive. A plain tower erected at Dover between 1180 and 1190 cost some 2.4 milliion ducats, at a time when crown income in England was probably no more than 12 million ducats a year. As stone was fairly cheap (one or two ducats per hundredweight), the principal element in the cost was labor, which is why construction was stretched over ten years.
If one wanted a more elaborate castle, and one wanted it in a hurry, costs would escalate rapidly. Thus, Chateau-Gaillard, the great bastion erected by England's Richard the Lionhearted (reigned 1189-1199) on the Seine above Rouen, was put up in a single year, 1197-1198, at a cost of 12,721,800 ducats, of which some 75% went for labor, representing some 2,544,436 man-days, the equivelent of over 6,000 men for a year. Materials, on the other hand, ran only about 2.4 million ducats, and transportation for such only about a million ducats more. When Edward I of England (reigned 1272-1307) bound Wales between 1277 and 1302 with a chain of ten of the greatest castles ever built, his total investment ran to something approaching 90 million ducats. Of course money expended on castles was money well spent, for they were long-term investments in military security. Castles were pretty much invinceable, for the art of siegecraft had not kept pace with that of fortification.
The best method of taking a castle was to use one's wits, to get inside by making a reasonable deal, or by means of a clever ruse, or a judicious bit of treachery. Failing that, one had no choice but to attempt a siege. A siege was neither a pleasant nor a cheap enterprise. One's army had to sit down outside the castle, blocking all entry and exit, and pound away at its defenses. Moats and ditches could be filled in with bundles of wood or earth; walls could be smashed with battering rams, undermined by tunnels, or bombarded by great catapult stones; the defenders could be picked off by archery, flames, plague, or hunger. The variety of machines available for such work was remarkable, some hold-overs from Roman times and some newly invented. Taking a castle this way required time and was costly in both money and men, for the enemy would not remain passive in the face of such threats. Moreover, an enemy army might march up and force one to give up the entire enterprise, with nothing to show for one's considerable investment in men, money, and time. Of course if one were in a hurry, one could attempt to storm a fortification.
Taking a place by storm could take a variety of forms. If tunnels or battering rams and catapults were successful in bringing down a portion of the wall, one could send one's troops up over the rubble in an assault. Or one could try the same thing by going over the walls using scaling ladders and mobile siege towers called "belfries". Then it all was a matter of desperate hand-to-hand sword and axe work, with a huge butcher's bill regardless of victor. In a storm, the quality of one's manpower mattered little, for knights had to fight on foot, clammering over rubble or up scaling ladders. This gave the defense a distinct advantage. Thus, a storm could be even more costly than a protracted siege.
There was another way to take a castle, one certain to work unless a relieving army came up, and one which was very economical of manpower, the blockade. To blockade a place meant literally to prevent all exit or entry. Doing this long enough would cause any castle to yield, for it would eventually run out of provisions. This could be done at little cost in lives. But starving a place out was expensive in time and money, for one might have to wait months, and one's own army had to be paid and fed whilst waiting.
Nevertheless, blockade did work. Chateau-Gaillard, virtually impregnable to conventional attack, succumbed on 6 March 1204, after an unusual winter blockade which had begun the previous September. To be sure, King Philip Augustus II of France (reigned 1180-1223) conducted something more than a passive blockade, for he erected siege works and successfully stormed the outer walls, but by the time of the final French assault the combined effects of combat, hunger, and disease had reduced the defenders to no more than 140 able-bodied men. Taking Chateau-Gaillard cost Philip Augustus a pretty penny, precisely how much no one knows, but it was easily several million ducats. This was money well spent, for not only did the castle control traffic on the Seine, it also covered Rouen, the capital of Normandy, then in English hands. Three months after Chateau-Gaillard fell, Rouen was French again, and the rest of Normandy soon followed. Obviously no one could afford such an enormous investment on a objective of lesser importance, so the castle was still a reasonable investment for anyone looking for a little security.
In the late 14th century cannon technology had matured to the point where effective cannon for sieges were available. Early cannon were cast by bell makers using much the same technology applied to casting bells. However the soft bronze used wore out quiclkly and reduced the force of the explosion. In 1375 the French built one of the early successful cannon. Four ironsmiths, eight assistants and one laborer built three forges in Caen and worked 2,300 pounds of iron. The gun was made from longitudinal strips of iron welded together and bound with 90 pounds of rope. It was reported that during the siege of Odruik in 1377, 200 pound stone balls were fired. Also in 1377, a canon was ordered that was capable of huring a 450 pound shot. At the Siege of Odruik, the Duke of Burgundy fielded 140 canon. It was reported by Froissart that earlier that year at the siege of Ardes, the French canon pierced the walls. This was the first record of a wall being pierced by cannon shot. Eventually the cannon helped bring about the downfall of smaller nobles who could not afford to construct "cannon resistant" fortifications.