OK people gather around for a story of shame and misery, of betrayal and treachery, of the hideous oppression of poor ship designers and the foul fiends who lead them. Weep at their plight, o people, for it was indeed shameful.
When the French Navy realized they had to replace Clemenceau and Foch back in the early 1980s, the Government was very reluctant to release the cash. They only agreed to the construction of two ships if they were exact replacements for the two carriers due for scrapping. This meant they were restricted to 27,000 tons normal, 32,000 tons full load and a length of 240 meters. Also, that they should carry a group of around 32 aircraft. That, by the way, made them look very much like an Essex class (which isn't surprising, the Foch and the Essex are very, very similar in design terms).
Length first. This was actually set by the available drydock. At that time, DCN had a stranglehold on French shipbuilding and required their facilities to be used. The maximum length that could be accommodated within their available drydocks was 260 meters. That immediately lead to a problem. Aircraft are much higher performance than they were in teh 1950s, they require bigger decks to operate from. They needed much more deck space so the ship had to be crowded into that available length. The real joke is that just across the port is the biggest drydock in the world, it was used to build the SS France pre-WW2. But, it couldn't be used, DCN didn't own it.
So, why not lengthen the drydock? Well, the problem was that at the landward side, the extension to the drydock was blocked by a toolshed. Why not move it? Well, the toolshed was owned by the Ministry of the Interior, the drydock was owned by DCN, part of the Ministry of Defense. Moving the toolshed should cost roughly US$100,000 - who should pay? Interior said Defense, Defense said Interior, they never agreed and the tool shed is still there. So was the length restriction of 260 meters.
So, to get around the problem, the designers adopted a solution by which an unusually wide flight deck was adopted. This lead to a rolling problem (its a matter of vertical movement, for the same degree of roll, a wide ship has a much greater vertical movement at its outer limits than a narrow ship so a wide ship has to roll less if its extreme vertical movement is to be within tolerable limits). In order to reduce roll to an acceptable level, the French had to include an elaborate computer-controlled anti-rolling system. This works well but its wasted weight and space, both of which were at a premium.
The restricted length gave another problem. The aircraft on board have to be accelerated to a specific speed in a specific distance. The catapults available couldn't do that. So the French designed a short, high-acceleration catapult. It worked OK but dummy tests showed the force transmitted to the pilot was very, very close to that which would break his neck. Since French pilots carried sidearms and were pounding on the door wanting to discuss this with the catapult deisgners, it was decide dto abandon the new catapults.
Instead, the French bought American C-13-3 catapults. Problem was that these were much longer than the French design. Now, some technicality. On the Pepe le Pu, the flight deck is the strength deck, this is what gives the hull girder its integral strength. Cutting two long slots in the strength deck severely compromises the strength deck and thus weakens the overall hull strength. Worse, because the design of the carrier was restricted in size, the two catapults couldn't be installed in the bow, there was only room for one there. The other had to be put in the waist. That meant not only were the slots cut in the strength deck long, they were one-behind-the-other and overlapped. That critically weakened the flight deck. The only option was to strengthen the flight deck by thickening it up and, because of its area, that cost a lot of weight.
At the other end of the deck, the French also designed a very fierce arrester wire system to bring teh aircraft to a halt quickly. It worked extremely well, the arrester wires stopped the back half of the aircraft perfectly. Unfortunately, the front half kept going. It was decided that this was not desirable (the pilots were pounding on the door again, this time with sidearms drawn) . The only option was to install a conventional arrester wire system and extend the angled deck forward. This interefered with the bow catapult and meant the carrier could not launch and recover aircraft simultaneously - a major limitation.
Later, it was found that they hadn't lengthened the angled deck enough, it was three meters too short and the heavier aircraft would still be moving forward when they reached the end and vanished over the edge. It had to be extended post-completion.
The travails of the design team weren't yet over. The problem now aorse of the elevators. They'd decided on two elevators, fair enough for the proposed air group. The problem was that the flight deck was the strength deck and stress levels there were already critical. Even deck edge elevators require a major cut-out in the deck and weaken the deck significantly. Now, the conventional solution is to put elevators on both sides of the ship, that's good from damage control and from weather shelter perspective. Unfortunately, that means there's a cut out on both sides of the ship, doubling the weakening effect. Having both elevators on the same side of the ship doesn't do that. So, the French had to put both elevators on the same side of the ship. The question was, which side? If they put them to port, they would obstruct the landing deck, they had to go to starboard, the same side as the Island. Hold that thought.
Now we need a little digression. For mostly political reasons it had been stipulated that the ship would need nuclear power. The problem was the French didn't have the money to develop reactors specifically for her, they had to use the K-15s off the shelf. Now, the K-15 was designed at a time when the French were hoping to export nuclear-powered attack submarines. To get around the nuclear non-proliferation treaties, they designed their reactors to use low-enrichment fuel called "caramel". This had two impacts, it lowered the power density of the reactor and it reduced the life between refuellings. Neither mattered too much in teh putative export SSN. It turned out they both mattered a lot in a CVN.
The French Navy adopted the K-15 for its own use. In theory at least, this isn't such a dumb decision. Caramel is around 15 percent enriched, the fuel used in US or British submarines is a lot more (like multiples) more enriched than that. However, highly enriched fuel needs special processing and reprocessing plans that caramel doesn't. So, in theory at least, its possible to argue that the financial and operational costs of using low enrichment fuel can be offset by the elimination of the reprocessing plant. Implicit in that approach (which turned out not to be the case by the way) is designing the ship so that refuelling the reactors is quick and easy.
This meant designing her so there was a path through the ship, straight down to the reactor room. For weight reasons, the reactor room has to be more or less amidships, so this path had to be amidships. Again, structural reasons meant that this non-load bearing path (essentially a hole cut right through the ship's girder) couldn't be in parallel with one of the elevators (stress levels in the flight deck would pass critical).
Back to the held thought on elevators. The refuelling soft patch had to be between the elevators, meaning the two elevators would have to be both starboard side and very well separated. The island (a structure that imposes significant loads on the ship) also couldn't be parallel with the reactor access soft patch. That meant it had to be either forward of the fore elevator or aft of the rear elevator. The latetr was impossible, it would have perched the island right on the stern.
The forward position had to be chosen. This is good for ship handling, lousy for aircraft operations. The French tried to claim that the forward position was selected because it sheltered the aircraft from the weather. The world laughed.
Still more problems. Flight deck space was critical, the minimum required was calculated and subtracting that from the space available gave the space for the island. It wasn't very much. All the antennas were compressed into a small area and they all interfere with each other. That means that many systems can't be operated simultaneously including such things as comms, search radar, fire control, radar et al.
More problems. The design chosen had a wide hull for its length. That's bad for speed. The French did a lot of research into hull forms, a lot of calculation and a lot of trials with a sub-scale model. None of which helped. Pepe le Pu was designed for 27.5 knots; it was admitted that there was no possibility of getting her past 25.5.
Other problems emerged as well. The reactors lacked steam capacity to operate the catapults properly (sustained launching of aircraft would deplete the steam capacity and have a serious effect on the ship - not just on speed). On trials that was a problem, but a bigger one was that the screws fell apart. Partly this was a production problem; the screws had been improperly cast and contained voids, but the other factor was intense vibration at higher speeds. After her screws fell apart, they had to be replaced by a set from one of the older carriers. They were unsuitable and restricted the ship's speed to "less than 23 knots" (actually 21.4). When her new screws arrive (2008), her speed will increase to above 23 knots (actually 23.5 her real maximum operational speed, she can do a bit more but the vibration is intense and its not recommended).
The ship had other operational problems as well, internal flow is not good, supply of munitions is difficult, all the things one expects of a new design team. By the time the design process was finished, the ship weighed 36,600 tons standard, 42,500 tons full load (35 percent and 32.8 percent overweight respectively)
She would have been a much better, more capable ship if she'd been designed as a dead dinosaur powered STOVL carrier but that would mean the French buying Harriers.