Note in your examples that images at
row 4, columns 5, 6 & 7 and
row 7, columns 7, 8, 9 &10
have a faint "white" bar at the top of the image.
Increasing the brightness and or gamma on the sample supplied shows what appears to be a reddish line source.
Doing the same on individual images may give you some visual clues as to what happened.

The source LOOKS like a light or heater.
If it's an IR source it may be affecting metering, but seems unlikely to feature in so many photos.

Did the samera show a black image immediately post exposure - even if not post viewing every shot it is offten enough useful to do an occasional check to pevent situations like this occuring.

Also, other 'black' images may have some image data present - trying adjusting a few.

Knowing how the camera was set and used may help.

Access to a few of the original images with EXIF data intact would probably help greatly with post mortem. If you have a Dropbox or similar account, making a few originals available may help. In this case IP loss is not liable to be a problem with full res images :-).

_____________________________________________

If a photo appears bad on "post view" in the camera it almost certainly is - cameras "break" only occasionally, but a bad image is ever only a change of settings away. So the chances of the camera displaying wrongly but of the actual image being OK are small. Therefore, checking results occasionally and adjusting the camera to get at least some sort of image is essentially essential.