Girder crown stays – are they overstressed?

[GoCreateParticipant @gocreate]

I have had some email correspondence with Paul regarding his original post, “Are my boiler stays over stressed”.

To answer his question I carried out an FEA (Finite element analysis) based on the drawings of the Durham traction engine boiler by John Haining. The results are shown below.

The materials are all copper except for the boiler firebox side stays which are specified as bronze.

The analysis confirms that, for the copper crown stays, the stresses at an operating pressure of 90 psi range between 6 and 24 N/mm2 well within the design limit of 26 N/mm2. (3770 psi). This stress is much lower than the 113 N/mm2 (16400 psi) which Paul calculated correctly for a simply supported beam.

The bronze stays have a maximum stress around 50 N/mm2, again well within the material yield strength.

 

This analysis verifies the original designer provided a very sound design which has been demonstrated by the many examples in steam for many years.

 

Nigel

Edited By tractionengine42 on 19/12/2011 12:52:30

[Sub MandrelParticipant @submandrel]

Hi Nigel,

 

Just a thought out of curiosity, but presumably that design would benefit from piercings in the crownstays to allow water to circulate over the crown.

Is it difficult top see what difference these might make?

 

Neil

[JasonBModerator @jasonb]

Hainings design does have cut outs in the bottom of the girders.

 

J

[Paul HorthParticipant @paulhorth66944]

I greatly appreciate Nigel’s work in carrying out the FEA analysis on the firebox girder stays. I think he has done the hobby a service.

It is now clear that the stresses in these girder stays designed by John Haining, and, by similarity, also the Martin Evans designs, are within the safe working limits used for other parts of the boiler. In my view we can conclude that the calculation of the girder stress using a simple beam model, (the point from which I started this thread) gives results that are quite unrealistic. The prediction of firebox collapse has not happened, and now we know why.

 

Nigel’s work is a vindication of the design talent of the two eminent model engineers, Evans and Haining, who had to work without FEA, but using their judgment and experience, backed up by experiment.

 

Paul

[GoCreateParticipant @gocreate]

Hi Stub/Jason

I have subsequently included the cutouts in the girder and cross members. There is no change to the stress levels only some slight changes to the stress distribution as you would expect.

 

It’s best to build up FEA in stages, then you learn more about the application and it makes it easier to resolve any FEA model problems.

 

Paul, this analysis is indeed a vindication of the original design work.

 

Seasons greetings and best wishes to everyone for 2013

 

Nigel

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Martin Evans ‘ Princess of Wales’ boiler design has four large sections of the actual stay lower flange cut away leaving only effectively a strip of the the top of the firebox itself as a working flange . This makes the crown stays much less effective and if it where not for the very high intrinsic strength of the firebox itself due to its narrowness and arched top the design might be be inadequate .

 

Are you actually saying that I’m wrong about this and that those enormous cutaways make no difference ??

 

I’ve put a bit of the drawing in my photo album .

[JasonBModerator @jasonb]

No one mentioned the cut aways in Martin Evens’ designs, we were talking about John Hainings D&NY traction engine boiler, that has only two cut outs not four and they are shallower at only 1/8″ high at the crest.

 

And as the Evens’ ones only take out 20% of the stay height its hardly an “enormous cutaway”  Its also a thicker firebox at 1/8″ compared to the D&NY at 3/32″ thick and the three gassets across the firebox crown are L shaped on Evans’ similar size boiler as opposed to hainings straight ones.  The stays are also twice the height of teh D&NY so again far nore effective eg there is still 1″ of metal left after cutouts as opposed to 1/2″ on teh Durham. So all in all probably the loco is less stressed than the traction engine boiler.

 

J

 

Edited By JasonB on 21/12/2011 16:51:17

[JasonBModerator @jasonb]

Michael, I have added a similar drawing of the D&NY firebox to my albums so you can see that teh stays are far less substantial than on the PofW. yet they still are well withing safe limits as shown by Nigels analysis.

 

J

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Jason ,

 

I think the general subject of these girder stays should be explored a bit more yet before any real conclusions are drawn .

 

The point about the cutaways going through the bottom flange is that a beam with a bottom flange is very strong whereas a beam with a cut away bottom flange is very much weaker . If it turns out that the stress levels are still acceptable then ok but it is still an absolutely awful bit of design work .

 

I’ll have a look at your drawing and get back on this .

 

MW

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Jason ,

 

Thanks for showing me the D&NY drawing .

 

The crown stays on this where obviously designed by a competent person and are probably as good as can be in a practical set up .

 

They are , however , radically different to the crown stays on Princess of Wales . I know its not obvious but doing stress analysis properly requires the examination of every minute detail of a component .

 

Regards the D&NY stays – main points only :

 

(a) Designer has left the bottom flange intact (as drawn) .

 

(b) Cutouts are relatively small , inside the flange and towards the ends leaving a strong section of undamaged beam in the highest stressed central section .

 

If Princess of Wales stays and D&Ny stays where scaled so as to be the same size then D&NY ones would be very much stronger than Princess of Wales ones .

 

MW

[JasonBModerator @jasonb]

I agree that the flange does add some strength to the part but the majority of teh load is being taken by the vertical leg of what is essentially an unequal angle particularly the upper part of it which is in compression, this can be seen in the analysis above as the lighter blue area, the flange being one of the most lightly stressed parts.

 

The flange plays less of a role in an angle than it does in a UB, UC or RSJ where the web is a thinner section and really only there to space the two flanfes apart. In an unequal angle the flange helps a lot to prevent sideways deflection of the main leg

 

How would you suggest the stays are fitted to the crown to avoid a build up of silt and allow watertion?

 

J

[Geoff RogersParticipant @geoffrogers81118]

I am not sure how many Princess of Wales boilers have been built, but I presume those that have have all passed the hydraulic test and are hence structurally ok. Presuming they have been built to drawing, that is.

 

Geoff

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hello again ,

 

The lower flange on the D&Ny design is only lightly stressed because it shares its load with a strip of the firebox roof – if the flange were not there or interrupted then essentially the same load would have to be taken by a strip of the firebox roof alone and that would certainly then be more stressed as a result .

 

The few full size boilers with girder crown stays seem to have taken a simplistic approach and just rivetted pieces of standard angle and channel back to back and to firebox . The only provision for water flow and cleaning seems to have been a few holes drilled on the neutral axes .

 

On the D&NY boiler the designer has probably done ok in putting in a few small circulation holes in relatively harmless places . I’ve not said that crown stays shouldn’t have holes in them but the sizing and positioning of these holes requires detailed thought .

 

The best general design solution for girder crown stays in models would be to make them as ‘proper’ beams with two channel sections back to back giving flanges top and bottom , no cutaways to the flanges and holes on n/a . These holes could be relatively quite big . If these were designed properly they would be immensly strong and that being the case could be made much lower in overall height with additional benefits for water flow and cleaning .

 

Regards ,

 

MW

 

 

 

[GoCreateParticipant @gocreate]

I have read with interest Jason’s and Michael’s comments above, both making valid points.

I quick look at the Princes of Wales boiler does show some interesting differences.

 

While I agree that the P of W girder stays look more substantial there arrangement/connection to the fire box crown has differences that will change the loading condition and stress distribution. Another observation is the Prince of Wales boiler appears to have a flat fire box crown (albeit thicker) whereas the D& NY has a curved fire box crown.

There is no question about the integrity of the design but I think it’s interesting to learn more about these girder stay designs. Michael, if you could give me more boiler details I would be interested in doing some analysis. If you send me a PM I will give you my email address.

How much does cutouts in the girder stays help water circulation? How much do they improve steaming?

 

Nigel

[Paul HorthParticipant @paulhorth66944]

On the question of cutouts in the girders, I have checked through my 1990s magazines for the more recent boiler designs by Martin Evans.

 

In the six designs below, he used girder stays without ANY cutouts.

 

Roedeer 1994

Eastern Belle 1996

Caledonia 1996 (7 1/4 g)

Adams T6 Flyer 1997

Harringay 1999

Greene Queen 1999 (7 1/4 g)

 

I no longer have magazines for the earlier designs so I can’t check those, but it would seem that at some point after the Princess of Wales(1971), Martin Evans decided that these cutouts were not worth making for improved circulation, and/or that they might weaken the stays.

 

In the Durham boiler, I did what has been suggested and simply drilled 1/4 inch holes in the vertical webs. This was less work than cutting the crescent shaped holes on the drawing. Nigel has confirmed that these cutouts do not affect the stress in the girders. I can’t judge whether these holes have much effect in improving circulation or reducing scale, but perhaps someone who has built one of these recent designs could have an opinion.

 

Paul

 

 

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Geoff :

 

Martin Evans on his original Simplex design argued that a successful three times working pressure test validated the design . This is simply not true – all it proved was that the boiler did not fail at three times working pressure which though useful information doesn’t actually get us very far in understanding what’s happening to the structure of the boiler under pressure . The three times pressure test effectively showed that the boiler has a safety factor of at least three BUT it does not say actually what it is – it may be 27 or it may be 3.001. Most model boiler codes require a safety factor of about 8 and to properly validate a boiler design one needs to know that this safety factor requirement is met in all sections of the boiler . There are only two ways to actually do this :

 

(1) By detailed calculation or

 

(2) By a type test where a representative boiler is tested progressively to eight times working pressure (and then scrapped ). This need not be done actually for each new design but it would be instructive to do it at least once when new design features are introduced .

 

Personally I would be happy to go with the calculated results and a normal boiler test but unfortunately there are very few people offering new designs with the training and experience needed to do the calculations properly .

 

 

I’m a bit sad to say (1) (2) above because many safe boilers have been built on a simple suck it and see basis over the years and there have been very few failures . But with modern day totally over the top boiler code and insurance requirements we actually have to KNOW that these boilers are safe .

 

Nigel :

 

I keep referring to Pricess of Wales because I looked at this boiler in great detail many years ago . Martin Evans dropped the cutouts in the stays in some later designs I believe because via a third party he heard of my concerns . Unfortunately he chose to simply remove the cutouts rather than have a reconsideration as to

where to put them .

 

The Princess of Wales firebox is narrow and has an arched top and these features save the design – it would actually pass a normal two times pressure test with no crown stays fitted at all !

 

I shall put my Email address in your pm and I shall shortly put the cross section view of PW firebox in my photo album – please ask if anything else needed .

 

All :

 

If I ever build another boiler it will have rod stays firebox crown to outer shell . There are so many if’s , but’s and maybe’s with these girder stays and so few advantages that I personally don’t think they should be used any more on new designs .

 

MW

[GoCreateParticipant @gocreate]

Hi

Micheal William’s interest in the Prince of Wales Fire box Girder Stays propted me to do some FEA on his behalf.

 

The results are as shown below. More results are shown in my album.

 

As can be seen the above is without girder stays.

and the above with girder stays.

 

These results show reduced stresses provided by the girder stays.

 

The analysis is done at a working pressure of 90 psi, material copper.

 

Even without the girder stays the design is quite robust IMHO, that’s not to suggest they should be omitted.

 

I hope these results are helpful to those designing and building boilers.

 

Enjoy

 

Nigel

 

[robert mortParticipant @robertmort83504]

Posted by tractionengine42 on 21/01/2012 20:06:17:

Even without the girder stays the design is quite robust IMHO, that's not to suggest they should be omitted.

Fascinating old thread, which should be reference on this subject which is hardly treated elsewhere.

Although the working pressure analysed above is 90psi showing the inner firebox without girder stays should survive, the usual test at 1.5 times this pressure would surely lead to collapse. So the stays provide the necessary safety margin.

[Author]

Posts