When high temperature, high pressure steam enters the cylinder, it comes into contact with the relatively cool walls. The wall surfaces are cooler than the steam partly because they lose heat by conduction and radiation, but more importantly they are cooled by the steam itself as it expands during the release and exhaust phases of the power cycle. Thus it may be deduced the temperature of these surfaces varies around an average that is a little lower than the average steam temperature over the cycle. As a resuheatlt, not only are these surfaces cooler than the incoming steam and so absorb heat from it, but they are warmer than the outgoing steam and so give up heat to it.
The fact that heat is transferred from the hot to the cylinder walls during the expansion phase and heat is transferred to the walls during the beginning of the compression phase, means that both these phases are not adiabatic. In consequence, the expansion index (n in the equation PVn = k) during these phases varies from a maximum (adiabatic) value of 1.3 to maybe 1.2 or in the case of short cut-off working and slow rotational speeds, even as low as 1.1.
Condensation: Because steam locomotives have traditionally been inadequately superheated, the term "wall effects" is often used synomymously with "condensation".
Condensation of steam inside the cylinder results in a massive waste of energy as described in detail in Porta's "compounding" paper published in Camden's book "Advanced Steam Locomotive Development - Three Technical Papers".
Porta describes wall effects as follows:
"Wall effect phenomena occur as follows. The cylinder cover, the steam passages and the piston (where the wall effects are most intense) are at a temperature between that of the live steam and the exhaust steam. Therefore, when the superheated steam makes contact with them, an energy drop takes place which show as a temperature drop. .... If the temperature of the walls is higher than the saturation temperature, no condensation occurs - the heat transfer being governed by "gas" laws, and is very small. But as shown by experimental data, if the temperature of the confining walls is below the saturation point at steamchest pressure, then condensation occurs even if the steam is highly superheated. This situation is the most frequent one in the life of steam locomotives: one of the causes is poor insulation, leading to heavy cooling down because of the intermittent nature of railway work.
The Author's measurements, even if rough, show that when the cylinders are well warned up after a lengthy, strenuous pull, in an ordinary locomotive in which the steam temperature attains 400°C, the walls reach a temperature higher than 210°C only at cut-offs greater than ~20%. Therefore, at shorter cut-offs, condensation occurs and the machine works with wall effects approaching those of a saturated one.
The temperature drop (in the case of no condensation, later to be defined) is roughly inversely proportional to the cut-off and to the rotational velocity to the power of -0.3 (ω-0.3). Thus, in the case of shunting locomotives, single expansion engines work with wall effects corresponding to saturated engines working at very low speeds, say ~1 rps, hence very high.
A large number of tests reported by GUTERMUTH show that condensation in saturated engines fed with steam at 8 to 12 bar, between points 6 and 2, Fig. 1, amounts to 40 to 50% of the steam admitted to the cylinder (and even more). And this refers to STATIONARY engines. What can be expected in the case of well "ventilated" [i.e. poorly insulated] locomotive cylinders?"
Fig 1 referred to in the text above is copied below with some corrections. It is also slightly simplified for clarity:
In this diagram, the variable steam inlet pressure between points 1 and 2 is replaced by a straight horizontal line G between 11 to 211 representing the mean pressure of the steam entering the cylinder. The location of this line is derived by equating the hatched areas above and below it.
If steam entering the cylinder comes into contact with surfaces that are below the saturation temperature for the steam, then condensation will occur onto those surfaces. Each cubic centimeter of condensate representing perhaps 1000 cc of useful steam (depending on its pressure). In any case it represents a contraction or loss of steam that is represented by the segment 21 to 211. This in turn results in an irrecoverable* loss of work or energy as represented by the hatched area H.
[* Porta points out that the condensate is likely to vapourize during the compression phase, thus recovering some energy that supplements the draught created by the exhaust system. This is offers no compensation for the loss of energy that might otherwise have been used for moving the locomotive and its train.]
In his "Compounding" paper, Porta goes on to say that:
"the basic principal concerning wall effects is to have at all times the temperature of the walls above the saturation point. It is therefore essential to have the highest possible steam temperature, even beyond the theoretical optimum. This point was missed by Churchward and all the British designers who followed him, with the possible exception of Gresley. The Americans also missed it.
The Author coined the expression "exaggerated insulation" ("hyper-exaggerated" for fireless locomotives) which implies the substitution of sealed-for-life fibreglass type material in place of cancer-causing, and less efficient, asbestos, and also the concept of insulating the whole cylinder front end, including the smokebox saddle, and its bottom, comprising the frame. Perhaps the most important point is that of keeping the cylinders hot between the working spells inherent to the intermittent nature of railway service. The "sealed-for-life" concept is a critical part of this; Wardale reported that in South Africa some 70% of the locomotives had no insulation at all!
A further attack on the problem could be the application of ceramic coatings on the surfaces: around 1890 Thurston (a very clever man) proposed painting these with a mixture of graphite and linseed oil. The Author tried it, but for trivial reasons the experiment was not followed up.
Another means of reducing wall effects to negligible proportions is to use steam jackets, but not those adopted by Chapelon for the 160A1 which required very complex castings. A welded fabrication is much simpler. These should be fitted to the cylinder covers, and around the steam passages, but NOT to the cylinder barrel. Jackets can also be fitted to the cylinders of fireless engines operating with saturated steam. Another difference from Chapelon's scheme is that the jacket condensation is not mixed up with the main steam flow, but re-injected to the hot water reservoir.Porta makes the point that condensation is equivalent to steam loss which necessarily increases steam consumption, and that this this directly reduces a locomotive power output as illustrated by his oft-repeated equation:
Porta concludes that wall effects or condensation can be minimised by:
- Using the highest possible superheat temperatures;
- Providing the best possible insulation of all external hot surfaces such as cylinders, cylinder covers, valves, steam chests and even the smokebox saddle to which steamchests and cylinders are connected;
- Providing steam heating to cylinder covers (but not cylinders);
- Using the smallest possible wheel diameter to maintain a high revolution rate;
- Using the smallest possible cylinder diameter and long stroke.
He also points out that:
- Short cut-off working results in lower surface temperatures and increased likelihood of condensation. Better to operate at >20% cut-off, illustrating one of the advantages of "compound" working;
- Condensation is inevitable during warm-up periods - usually around 20 minutes of operation on full power. Shunting locomotives are likely to suffer from large condensation losses.
- Fireless locomotives operating on saturated steam are also likely to insuffer from large condensation losses.
Several pages of this website include text and diagrams copied from Porta's "compounding" paper, including the pages covering steam leakage, clearance volume, incomplete expansion and triangular losses. More specific references to his theories on compound expansion can be found on the α Coefficient and Compound Expansion pages.
Sincere thanks to Adam Harris of Camden Miniature Steam, publishers of "Advanced Steam Locomotive Development - Three Technical Papers" for allowing the sections of the book to be published on this website.