Boiler design & construction ex Patriot thread.

If it is a new boiler, even to an old design, it needs to comply with the Pressure Equipment Regulations. This effectively requires approval by a 'notified body'. There is an official list of notified bodies (Google it, if you want!) but generally you are talking about various Engineering Inspection companies, such as R&SA, Zurich, etc. Have a look at: https://www.zurich.co.uk/internet/home/SiteCollectionDocuments/Engineering/710352001LoRes.pdf
If it is an old boiler being repaired, there is no requirement to comply with the PER, which is why many prefer to repair even the most desperate of boilers. With this, the need is to comply with the Pressure Systems Regulations which effectively require over-inspection by a Competent Person (Boiler Inspector), who will ensure that the repairs are to a satisfactory standard.

 

I see PER is actually a comprehensive document specifying the design and manufacturing criteria, including the use of proof testing to determine acceptance.

This explains a great deal. Thank you very much indeed.

 

Before the construction of the boiler started we provided copies of all the original Crewe boiler drawings to our insurance company for an assessment of the design to be made. The result of this assessment was that the design was sound and probably over engineered for the job. We were told that it is suitable for operation at 250psi, and we will build and test the boiler to this pressure, but it will only run in the locomotive at the original 200 psi. The only changes have been to weld the barrel seams, rather than use a riveted butt strip and a recalculation of the size of firebox stays.

With regard to drilling to firebox stay holes in the wrapper, the firebox is assembled complete with both Inner and Outer boxes before the stay holes are drilled to ensure the holes line up.

 

Did you mean to put the 'in' in that sentence Kev ? I wonder whether the fact that the Patriot boiler is over engineered, will be of interest or maybe of relevance to the County Project also, but maybe not for this thread !

 

Thanks for this.

If you are welding the Barrel, are you tempted to weld any other sections?

Do you know if the stays are going to be a larger or smaller diameter?

 

Very interesting stuff. I find the comment about being over-engineered especially of interest. If you design a land based industrial boiler the conditions are pretty static and known so sizes can be calculated. Such are the principles upon which BS2790 are based and on which the analysis will probably have been based, because you have to assess against something. However, steam loco boilers have additional stresses and strains applied to them, the majority of which are not determinable. Steam locomotive designers largely based their design detail on what they knew worked. it was a compromise of weight against durability. To keep the weight down they would look to reduced thicknesses and sizes whilst keeping to the basic calculations. If that gave trouble, they would make it bigger until a satisfactory compromise was reached. Having a welded barrel means that, in theory, you can have a thinner one but then your corrosion allowance becomes less and modern boiler design does not have much of a corrosion allowance. Locomotive boilers are always compromise in their design.

 

In theory, one could explore the use of more corrosion resistant alloys (steel based) if you were going to weld the barrel, and indeed potentially slim down the thickness, as they are all thin walled anyway.

I beg to differ. Modern methods allow nearly all stresses and strains to be determinable. The beauty of Finite Element Analysis. http://www.consuta.org.uk/workshop/FEA.html

Furthermore, these techniques are being used to take another look at designs to implement improvements.

One example being the finding that a radial stay pattern may be more suitable and has the potential to reduce severe stresses found in throat plates and foundation rings, which arise because of distortions in the firebox inner wrapper. http://www.consuta.org.uk/workshop/images/New Boiler/Boiler 10-4 Drawing.pdf

If you follow the links, you will actually get the cyclical animations of the boiler being brought to pressure and back down again to show how the stresses travel and distribute and the strains they cause.


.



Here in this last one, on the right is the traditional vertial/horizontal stay pattern with a 1.6MPA (230psi) pressure. By changing the stay geometry to a radial pattern, throat plate and foundation ring stresses were significantly reduced. This arises primarily due to the higher overall rigidity of the firebox.

 

All very interesting and well worth having a good look. Radial staying of fireboxes is a fundamental of the Churchward/Player firebox and has been well understood for many years, if not due to FA. Very few inner fireboxes have flat crowns, as shown in your boiler drawing. Finite element analysis may well allow better understanding and design of boilers. Does your analysis include for the effects of temperature and expansion or is it purely looking at stress and strain?
Putting aside corrosion, the problem areas of boilers are fairly well known. Stay breakage due to fatigue failure and necking is common, as is grooving of steel plates, especially at foundation ring and smokebox tubeplate. Star cracking of stay holes is another common failing and cracking of firebox shoulders is common on larger boilers. Loco boilers rarely fail due to plate failure, unless they have single lap joints and consequent grooving. Peaking of welded boilers is a modern day equivalent problem. .Many other problems are not stress related; leakage of stays and tubes being common ones, fireside erosion is another, which can be due to ash erosion and plate burning. The latter can be significantly affected by scale on the water side, which is one reason why good, well controlled water treatment is imperative.
My reference to indeterminate forces relates to those due to the fact that a locomotive is a very dynamic object, with racking and acceleration forces all all having an effect on the boiler. Such things will not apply in the case of a land based boiler or your launch boiler.

 

I must stress these are not my own models, although I do produce similar work.

Thermal effects can easily be incorporated in to the analysis, however, I think these models are assuming homogeneous material constants and assuming the primary stresses are Hydrostatic.

Stress, strain, creep, fatigue are all areas one can analyse among others. As we can see, the stay anchor points are always heavily stressed due to the nature of the design. there is little getting away from the problem, aside from upping the stay diameter to overkill levels. As you said, compromises.

Are most inner fireboxes of a semicircular shape? I always thought you had a square inner with an either square or semi-circular outer? How are the stays fixed down with holes coming through the inner box at odd angles?

I did think you might have meant that but I wasn't sure. However again we could apply some track data, same method as the Vampire models gathered by Tornado and applied to the P2, and add these to the analysis with suitable constraints to look at the dynamic loading effects on the boiler.

 

The majority of problems in locomotive boilers are caused by thermal expansion. Any analysis must take this fundamental fact into account.
Here's a section through a N.E.R T2 (LNER Q6) firebox, which clearly shows the curved firebox top and radial staying. another fundamental is the increasing waterspace dimension with height above foundation ring, which also allows increasing stay length to give greater flexibility. In this respect, increasing stay size (diameter) increases rigidity, which in itself, leads to stay breakage problems due to higher bending stresses. Stays really need to be long and (relatively) thin because flexibility is paramount, especially in the traditional breaking zones..

It is interesting to note that your FA shows that the outer firebox wrapper is overloaded. In my experience, this is one area of the boiler which never gives any trouble. Any problems with crown stays and their interface with platework is generally due to necking of the stay immediately above the firebox crown. A prime example is this:

I can't see any analysis method that would eliminate such problems.

 

Yes, I agree, thermal issues must be considered as only taking hydrostatic effects does neglect some pertinent facts.

I see what you mean now, there is a slight arc to the inner firebox to make the interfaces with the stays neater.

Are all the rows of stays held in the radius'ed anchor at the underside of the wrapper like that? Quite a clever way of solving the stay interface issue and increasing structural rigidity in the outer wrapper.

The outer wrapper is being pulled down by the hydrostatic pressure exerted on the crown of the inner box. However, the two move in tandem given the similar geometry, and so the stays move in tandem too.

So the necking is occurring on the stay immediately underneath the fire at the interface with the water and inner wrapper... That is some pretty advanced corrosion, is this driven by cavitation corrosion given this is where the boiling of the water is going on?

Just from looking at those models, I'm going to take a big risk here and nail my colors to the mast. I am of the thinking that the more rigid the overall design, that is the relationship of the inner and outer firebox, the less stress the stays have to endure because of reduced deformations. Deformations in any one area have a habit of having knock on effects in other areas, and by eliminating deformations, we eliminate these effects. That appears to be what the models show in cyclical loading, and is somewhat borne out by the fact that the designers have put in more and more stays as pressure increased.

 

There are various methods of staying when it comes to crown stays. This is just one of them. With most stays, flexibility is the key but it can sometimes be impossible to provide this. Having pin joints is generally good. With small fireboxes (and I would put your launch in this category) it is quite common to provide girder stays to support the firebox crown.
See http://www.svr-engineering.co.uk/boilershop/new_build_contruction.html or Google girder stays. This allows the outer wrapper to act as a circular vessel and is self supporting.

As I said previously, the outer wrapper is seldom a source of problems. The radius of the crown sheet can be a problem, though, and it is usually a function of the radius; the bigger, the better.

I don't think the phenomenon of necking is fully understood. My own pet theory is that, the relatively flexible stay is attached to the relatively rigid crown. The steel stay gets a layer of corrosion which, to some extent protects the parent metal. However, the constant flexing at the interface breaks down this layer, allowing more corrosion to occur and this effectively reduces the parent metal. it is a similar process with grooving at the foundation ring. With a steel/copper interface, electrolytic action also has an effect.

That is probably the opposite of the generally held theories, which look to stays, especially, being as flexible as possible to minimise bending stresses. This is all generally related to thermal expansion. However, at the same time as Churchward was evolving his boiler principles, Player was evolving a very similar boiler in the US. Churchward claimed that his design gave his boilers greater flexibility compared with existing designs. Conversely, Player claimed that his virtually identical design gave his boilers much greater rigidity compared with existing designs!
I think I have a comparison section between Churchward and Player boilers somewhere and, if I can find it, I will upload it.

edit: Found it. It is actually from the RCTS book 'Raising Steam on the LMS' which is a mine of information on locomotive boiler design.

 

A beam in bending to prevent the crown collapsing and allowing the outer wrapper to act as a shell vessel. Pretty neat!

Indeed, I see that from the drawings. That's not surprising the more I think about it, given the respectively larger geometry of the outer wrapper.

That runs contrary to my engineering education so far. The flat(ish) plates being the flexible items and the stays by comparison being rigid given they are predominantly in compression or tension depending on location and the plates being bent and sheared due to the surface loading. I was about to suggest corrosion resistant stays, maybe inconel, but this presents its own issue of cost and dissimilar expansion coefficients. The necking, in my opinion, is a direct result of loss of material due to corrosion, per standard tensile necking phenomenon. Stresses can be approximated through the stay as homogeneous, and any thinner sections will neck due to loss of area.

I'm not sure where the stays are in bending, I predominantly see them as being in a mix of compression and tension. The former potentially leading to localized buckling in worst case scenarios.

Sorry, which one is the Player, and which one the Churchward?


I fear we are getting a little off topic here...

 

Very much off topic, but interesting; at least for me.

The girder stay is very common on small fireboxes and is dealt with in BS2790.

I can assure you that bending stresses are far more significant that direct stresses in stays. The firebox expands significantly more than the outer parts of the boiler due to being in contact with the flame/hot gases at up to 1200 degC. On large copper fireboxes the differential expansion can be significant but is there, even on small fireboxes. The foundation ring holds the two rigidly together at the base so stays near there aren't subject to bending, as does the mouthpiece (firehole). As you get nearer the top of the firebox the difference due to expansion becomes significant and stays here are subject to significant bending. In boiler terminology, the areas near the front and back corners are referred to as the breaking zones and on large fireboxes, flexible stays were fitted to reduce the problem. However, these flexible stays aren't truly flexible, being rigidly fixed at the firebox end.

As to which is the Player boiler, I have to admit I don't know. The original drawing in the book does not label either. It's prime purpose, though, is to show the similarity between two boilers being designed several thousand miles apart by two different people. No doubt someone with better knowledge of GWR boilers can identify which is the Churchward boiler.

 

Fascinating, many thanks both, for a very readable and understandable MIC in Boiler design and performance, it has enhanced,what was, a very basic grasp of the subject and has pulled a number of seperate threads together to give a fuller picture.

Andy

 

The caption says a) is Player and b) is Churchward. Should be from left to right!
I would like to suggest to split this off into a boiler design thread, especially given the universal fear of
welded locomotive boilers, finite element analyses might open some eyes!
Kind regards
Jos Koopmans

 

Deleted

 

Apologies to the Moderators for the thread drift.

Tornado has all welded construction does she not? I'm not sure how trouble free that particular design has been, but I would personally rate an all welded design highly comparable and in some respects more favorable than a riveted one, primarily from a manufacturing standpoint. Additionally, I think there may be some room to do some very clever things with a welded boiler that might be particularly difficult with riveted ones. But I am as ever open to enlightenment on this point.

The best thing about FEA has to be that a toddler can visualize the problem areas even if they don't understand discretization or surface constraints. It is very intuitive.

Given the deep encyclopedic knowledge of members here and the wider preservation movement concerning working practice and common faults, I hold a belief that designs can be tweaked with the aid of modern analysis to preserve steam working for many decades to come.

 

I'm also very much a fan of welded boilers with steel fireboxes but I suspect that I'm in a minority in the world of heritage railways. Welded boilers have been commonplace for years but there are only a few welded locomotive boilers and they are generally relatively small. Tornado's boiler is the only large one and hasn't exactly been trouble free but large boilers are always more problematic than small ones. Again, this is largely a question of expansion, which is a function of size.

It would be interesting to see an FEA which took account of thermal expansion and practical plate temperatures.