If you look at the Nagaoka curves for the quantity of heat transferred through a material such as a metal sheet or shape (eg firebox) to a liquid, you can see that as the liquid boils, a sheath of vapour forms around the metal surface, which tends to insulate the metal from the liquid. Thus for a given heat input, the heat transfer goes down and the temperature of the metal wall goes up. This can go so far as to lead to melting of the metal. The Nagaoka curves show this by the temperature v. heat transfer graph increasing to a point of inflexion and then the gradient goes negative - it's basically a straight line to that point as I remember. I can see that items such as plate and rod stays will modify this, but has anyone any information as to how much heat transfer can be achieved through a firebox before boiling becomes so intense as to cause trouble? can it make the inside of the firebox more likely to suffer cinder cutting or other ills by tending to soften it? An extension of this, is of course, the effects of the build up scale , which is a more extreme case. Peter Chadwick Stockholm MODEL ENGINEERING DISCUSSION LIST. To UNSUBSCRIBE from this list, send a blank email to, modeleng-request@xxxxxxxxxxxxx with the word "unsubscribe" in the subject line.