You ask really good questions!

I really haven't done an significant amount of work with composites, but having seen damage to CF structures, it's hard to imagine how it's patched up as a laymen, especially if the damage is across a larger area. Must be a lot of skilled labour going into it!

The Wikipedia article on composite repair is pretty good, and gives some diagrams explaining typical repairs. Worth a look to understand how it's done.

https://en.wikipedia.org/wiki/Composite_repair

Say, how do you determine if the layers of the original composite part bonded correctly with the patched up part? Isn't this always a bit of a concern?

You're right that the bonded joint is the critical area in any composite structure.

Most of this is controlled by process. You remove the old material in a particular way. Then you clean it in a particular way. Then you treat the bonding surface in a particular way (roughness, cleaning, etching chemicals, bonding promoters, etc). Then you apply the adhesive in a particular way. Then you apply the layers in a particular way. Then you do a particular heating cycle for the curing process. All of those processes are developed and proven through lab tests, so we know that, if done correctly, they will produce the required strength (with a safety factor). Then at the end, typically a non-destructive inspection method (typically Ultrasound) is used to ensure that there are no voids (bubbles) in the bonded areas, and that any which are present are within limits (maximum individual size, minimum distance from edges, minimum distance from other voids, maximum total number and area of voids). If you do all that right, it's highly probable that your repair will have enough strength and will last long enough.

I did a series of bonded repairs once where we didn't have those highly defined processes, but it was also not as critical an area, and we didn't need a lot of strength. But what we did is had the technicians make a series of samples using the same surface treatment process and the same batch of adhesive and the same cure cycle with every repair they did. We then sent those samples to a lab where they were tested destructively for strength and durability. It turned out that, with a couple of outliers where the technicians clearly did not follow the process, the bond was 5-6 times stronger than we needed it to be. So, success. Plus, we had data to support an allowable strength for other repairs using that process in future.

I didn't know that either. I thought they would just remove the individual sections of the hull, frames and ribs and replace them with new parts.

It's usually a very big deal to replace a significant structural component on an aircraft. For example, replacing an entire fuselage skin section is a massive, massive repair. Instead, engineers will design repairs to patch, splice, or whatever to the area where the damage is located. The splices or patches will be installed with rivets or bolts typically, and usually they'll be in place for the life of the aircraft. The exact process depends on the structure in the area of the damage.

Another question that just came to my mind: How long are the longest continuous fibres in a fuselage usually? Doesn't that play a major role on how strong/stiff a composite part is? It's hard to imagine how to get the patch bonded in a way that it's actually load bearing and distributes the load to the interrupted fibre strands.

So as mentioned earlier, joints are usually the weak points in composite structures. So, the fewer the joints, the stronger the structure. How long is the longest strand? Well that depends on the part. For a while, continuous fibre winding was the rage, where they would effectively put the fuselage on a big rotisserie and wrap continuous fibres all around the thing and the entire structure was essentially made of a single fibre. That fell out of favor as a manufacturing method because its expensive and we can get pretty close to that using other means.

On something like a 787, I would imagine there are fibres that, for all intents and purposes, wrap all the way around the fuselage, but I'm not familiar enough with its construction methods to know for sure.

The thing is, if you imagine gluing two pieces of paper together with an overlap, you'll find that, once you get a certain distance from that overlap, the strength of the paper is no longer affected by the overlap. Composite structures are the same.

The needed strength comes from (and this is greatly simplified) how far you overlap the original layers with the repair layers, and the strength of your adhesive. So if you need the joint to carry 1000 lbs through it, and your adhesive provides 500 lbs of strength per inch of bond, it means you have to overlap by 2 inches to get the 1000 lbs of load transfered.

The bonus of composite structures is that the fibres are all very close to each other. Like, you're often trying to get a bond line thickness of 0.010 inch or something like that. So the fibres from the original structure and the fibres from the repair structure are really close together and between them is a very high strength vinyl or epoxy adhesive that will transfer load very effectively.

Oh, one last thing I'd love to know: Is it true that composite airframes last longer than metal ones on average? I've read that micro-cracks are a bigger problem in metal aircraft than they are in composite ones - is this correct?

Composite structures are not as prone to fatigue damage. So for them to degrade structurally, they require a whole lot more stress cycles. In some cases, you can assume its infinite. But there are other degradation modes, like water ingress (yes, even something like epoxy can, over time, wick moisture through it into the structure), UV resin degradation, barely visible impact damage (if you drop a screwdriver on a composite structure, you can cause invisible disbonds between layers of the structure). Hail damage, stone chips, bird strikes, all require much more extensive investigation if the structure is composite.

Metal structures are easier to repair, simpler to manufacture in many ways, less prone to many of those threats (though corrosion is an ever-present problem), but does tend to crack over time. But those cracks can be predictable and repairable. They are also heavier (typically) than an equivalent composite structure.

So yeah, generally composite structures will last longer before developing fatigue problems, but may instead develop basically "rot" problems or impact damage problems instead.