The Freyssinet hinges are from my viewpoint a device never to be used anymore. I think they have been responsible of fatal casualties in the ruin of a hotel in Málaga in in the 70's.

Upon creep of concrete, the steel may really be yielded at work status. This is coherent with the cracks you see in the hinges. This makes the joint less than ideal to refrain any force orthogonal to the main compression stress.

If a find reference on computation will post.

A two pages treatment on Freyssinet hinges I myself find in
(in spanish)

Hormigón Armado
Pedro Jiménez Montoya, Álvaro García Meseguer, Fernando Morán Cabré
7º Ed.
Ed. Gustavo Gili Barcelona 1975
p. 504

It is a classical RC spanish text and may well be in the over XV edition and still retain the Fr. hinges information.

If you can't get better thing, you could state your concrete questions and I would try to answer to the light of the text.

It is clear however that the problems you are asked to look at are compounded with and confirmatory of my opinion on lack of fitness for use of these devices, and almost any doubt I would see good their replacement for better understood support devices.

The more I look at these things, the more I tend to agree that using freyssinet hinges was an unfortunate engineering fad.  Unfortunately I am not in a position to replace the large number of bridges using these hinges and so I am looking for a way of evaluating them.
Thanks for the reference, I will try to get a Spanish speaker here to lead me through the text before I bother you too much.
Regarding the visible joint cracking, the vast majority of them only show a single hairline (0-0.5mm) crack at the soffit.  No vertical displacement is visible.  A very few hinges exhibit single large cracks up to 3.0mm wide with up to 1.0mm drop.  These hinges have been propped as yielding has been assumed.
A now demolished bridge of this design was strain-gauged a couple of decades ago with the conclusion that temperature effects would lead to dowel bars yielding at concrete temperatures below -15 Centigrade.  It is extremely rare for such temperatures to be reached in this country.  
It is possible that dowel bar debonding (possibly due to corrosion) is increasing the length under maximum tension and thus the bar stays in the elastic zone.  I am reasonably confident that the majority of hinges are not operating near yield.  
On the other hand, it appears fairly common that the dowel bars can be misaligned up to 100mm, increasing the axial load on the dowels.
Acoustic emission testing is detecting concrete cracking that is not visible at the soffit.  I am hoping to find out where the cracking is when I break out the 'noisy' joints to confirm the radiography section loss.  The joints previously broken out by others have not had any unusual cracking noted.  The amount of cracking detected by acoustic emission seems quite large while nothing is visible externally.  It is roughly equivalent to 60mm2 of mortar cracking per day.
One guess I have involves progressive concrete cracking from the dowel bars bearing onto the edge of the hinge.  Another guess is that the cracking is from dowel bars debonding along their length, assisted by corrosion.  Any other ideas?

Both

Theory and Design of Bridges
Xanthakos
p. 993

and

Structural Engineering Handbook 4th ed.
Gaylord, Gaylord, Stallmeyer
p.17

have something on Freyssinet, Mesnager and Considére RC hinges, but not much.

DaveMinter,

"Bridge Bearings and Expansion Joints" by David J. Lee has some info on Freyssinet hinges.


HTH

Thank you ishvaaag and Ingenuity,

I checked the references you suggested, they have proved useful.  I now have a fighting chance of actually understanding how these things work in practice.

It is interesting that everybody I've spoken to at work (including quite highbrow types) refers to these things as Freyssinet hinges.  They appear to be an unusual variation on Mesnager hinges.

Now I must put my head down to review the assessment calculations, prior to breakout.  Just pity me