Behaviour of doors exposed to fire

[kurnal]

I came across some beautiful historic  doors in a listed building the other day. Mahogany 50mm stiles and rails with fielded panels which appear to be made up of a 15mm core of softwood butt jointed planks covered either side with 2-15mm mahogany fielded cover held in place by beadings pinned in position. Someone in the past has "upgraded" the doors by installing  intumescent seals.

The walls of the building are curved and those doors and frames that sit in the curved walls are also curved. Quite markedly so- the bow on the head of the door frame is of the order of 60mm.

Now heres the question.  I am curious to know if any fire tests have been undertaken on curved doors and whether the the curvature is likely to accentuate the distortion caused by exposure of one side of the door to fire. My gut feeling is that pi will come into the equation somewhere and that any distortion of the line of the circumference is likely to be amplified at the edges of the door leaf but would like to know if this gut feeling is justified.  And whether it makes a difference as to which side of the door is exposed to fire.

(No worries about the doors in question by the way, I am recommending other protective measures to ensure they should never be exposed to anything other than smoke)

[Phoenix]

Hi kurnal,

No experience of this but I can offer some brief lay observations.  Curved timber can be of two forms, it can be cut or carved into the curve or it can be bent from straight timber, usually with the aid of steam. 

If the timber that forms the doors has been bent then heat from a fire would be likely to make the timber try to return to its unbent state. 

If the timber has been cut into a curve then there will be no overwhelming tendency for the timber to straighten but the face exposed to the heat may be marginally smaller or larger than if the door was straight (depending upon which face is exposed to the fire) but the difference really will be marginal and will undoubtedly be outweighed by other variables such as the quality of the door construction and the quality of the fit in the frame.

Maybe there is some scientific model for the performance of such doors but making it rigorous would be difficult.

Stu

[Fishy]

As Stu says, the way in which the curved rails are formed will have a very significant effect on its reaction to the fire.  I wouldn't imagine that there's any particular reason why a piece of timber machined into a curve would bow any more or less than a straight piece, so long as they were cut from the log in the same way (quarter-sawn, half-sawn etc). I agree with Stu about the steamed timber.  It could conceivably be laminated rails, too, in which case if the glue melts it'll all fall apart!  Intumescent strips could also make the distortion worse, if they are high pressure types, although at 50mm the rails may have enough strength to resist the bowing tendency, for a reasonable period of time.

I don't know of any models that predict the performance of timber doors when exposed to fire from 'first principles', but there are well-established techniques that extrapolate the performance of tested constructions (e.g. increased height, width etc...).