Why does my printed laptop shell keep cracking at the hinge after a week?

The layer lines are exactly where your crack should start if the hinge boss is printed flat, so your diagnosis is probaly right. When the screen gets yanked open, the load on the hinge puts the part in bending, and a flat-printed boss means the bending force is trying to peel the layers apart right at the top of the boss. PETG layer adhesion is decent but it's still the weak plane, and repeated opening and closing is a fatigue situation, so it only takes a week or two to find it.

You could try redesigning the hinge connection so it doesn't have that sharp 90 degree corner right at the attachment point. Stress tends to concentrate way too much in tight corners, no matter how strong the plastic is. If you soften the curve where the boss meets the main chassis, it should help distribute the load better across the surface instead of forcing it all onto a single layer line. It might sound simple, but those little geometric tweaks make a huge difference on a print like this.

I tried reducing the solid parts inside the hinge piece to make it lighter since I thought that would help. It didn't change the breaking, but the way it tore was diferent. It felt more like the structure was collapsing inside than the plastic shell splitting. I wonder if using a different kind of internal fill helps keep the pieces togather when they are moving back and forth. What kind of infill pattern works best for things that move a lot?

It sounds like you need more material around that stress point, especally where the hinge connects. When I design structural brackets or heavy components, I always run the wall count up to 5 per perimeter instead of the typical 3. That extra shell count adds rigidity and significantly handles the shear force better than just adjusting the print angle. Have you considered putting a small rubber gasket inside the hinge pocket? Absorbing that impact energy might stop the fatigue before it even starts.

Honestly, even with all the suggestions about wall count and rounding corners, I still don't trust the print for that kind of leverage. When I designed a heavy bracket last year, I stopped relying on the plastic integrity altogether for the load. I just drilled a hole through the hinge boss and used a small threaded rod and locking nut. It takes some extra machining time, but it completley isolates the shear stress away from the layer lines. If you're going to be using this thing regularly, you want a true mechanical connection rather than trusting the plastic's strength under flex.