Boiler Design – issue 4765

[JasonBModerator @jasonb]

Does the small size of our tubes start to cause cleaning issues once filled with a spiral? And our chimneys don’t offer as much draft against the now restricted smaller cross section of less tubes.

On a separate note while looking at a few of the usual suspects for boiler size tube 2.5-6″ it seems 13g has almost gone off the market so even if you caclate a thickness near that including the other variables you are going to have to go for 10g or roll your own it will end up as thick with a butt strip allowance added.

[Martin Johnson 1]

On Martin Johnson 1 Said:

When I was designing shafts in pursuit of money we used the Von Mises failure criterion which required the torsional and bending stresses to be combined basically by summing the squares and rooting the result (from distant memory).  I accept that the choice of appropriate failure criterion in this case could be quite a complex subject.  Anyway my point stands that it was not addressed in the original article.

Martin, I missed this follow up to your post #801179. Yes, you would do that kind of calculation in a situation where there are compound stresses such as a combination of torsion and bending.

In a circular thin shell under internal pressure it is just different amounts of tension in different directions, not a compound stress. The axial stress is half the hoop stress. And, qualitatively, at some angle to the axis, it is still just tension, but more than the axial tension and less than the hoop.

[MEinThailandParticipant @meinthailand]

The UTS Method Explained – Part 1 – Close Light

A Very Funny (But True) Story About The Thai Language And How It Relates To The UTS Method

I have been living in Thailand for the last 10 yeas and learned a bit of the language.

‘Pert Nam’ = (Open Water)
Pert = Open, Nam = Water.

So pert nam means open water. Let the water come out. They never say open the valve. Just open water.

‘Pit Nam’ = (Close Water)
Pit = Close, Nam = Water.

So pit nam means close water. Shut the water off. They never say close the valve. Just close water.

Similar expressions are used relating to electricity and light bulbs.

‘Pert Fai’ = (Open Light)
Pert = Open, Fai – Light.

So pert fai means open light. Switch the light on. They never say switch the light on. Just open light.

‘Pit Fai’ = (Close Light)
Pit = Close, Fai = Light.

So pit fai means close light. Shut the light off. They never say switch the light off. Just close light.

In my big workshop I have a 3 phase distribution board with (of course) an isolator or as the Thai say a ‘cut-out’. It isolated the DB and hence the workshop. I always switch off all the lights and isolate the DB before locking up at night and my Thai wife always asks me if I’ve done that in case I might have forgotten.

Two days ago after I had locked up for the night she asked the usual question “Pit cut-out mai?”. Mai just indicates a question and as we’ve seen above, Pit means to close, hence; “Close cut-out mai?” intending to mean did I switch of the power at the cut off? But literally it means “Close cut-out?”

Feeling a bit mischievous I said “No, I opened the cut-out”.

Of course she got angry and demanded that I unlock the workshop and “Pit cut-out” – close the cut-out.

As I tried to explain that the Thais have it all wrong in relation to switches and electricity she got even more agitated and angry. I tried to explain that you have open an (electrical) circuit at the switch or cut-out in order to stop the electricity flowing. In order to “pit fai” (close light) you have to pert (open) (not pit – close) the cut-out.

I refused to reopen the workshop and “close cut-out” and instead tried to explain that in order to switch off the power you have to OPEN the cut-out. She wouldn’t have it. And a flaming row erupted. She insisted that I “close cut-out”.

What has this got to do with the UTS Method for boiler design?

It just shows how hard it is get someone to change their way at looking for something when they’ve always done (or said) it that way for years. When it’s become engrained over many years it’s very difficult to show people a new way of looking at things. It’s just human nature.

So I’ll just say “Pert Dtaa”. Pert = Open, Dtaa = Eye.

I’ll try to do that in the next few posts.

Thank you.

[MEinThailandParticipant @meinthailand]

The UTS Method Explained – Part 2 – Fudge Factors

The UTS method takes the Ultimate Tensile Strength (UTS) of un annealed (i.e. hard) copper and multiplies it by a so-called “Safety Factor” to arrive at a safe allowable working stress for the copper.

This is completely fallacious as I’ll prove below.

Please note that some of the information below is borrowed from the article “The Yield Point Method ” by Les Smith and Alan Brown, published in Model Engineers & Workshop magazine in Volume 134, Issue 4765, June 2025 edition, page 26.

1. The Properties Of The Two Materials In Question; Hard Copper (UTS) Method and Annealed Copper (YPM) Are Totally Different.

Just look at the Stress/Strain curves for Cold Drawn and Annealed copper shown below:-

Clearly the mechanical properties for the two grades of copper are totally different. They might as well be different materials – as indeed they are – the only common denominator being in the name – copper.

But the UTS method takes the value of the UTS for hard copper and multiplies it by a so-called “safety factor” to arrive at a safe allowable working stress for annealed copper. Where is the engineering logic to take a mechanical property from one material and apply it to a different material?

These are the ‘Safety Factors” (SF) quoted:-
Example 1:- SF = 8
Example 2:- SF = 8

In his book Boilers and Boiler Making KN Harris quotes a Safety Factors of 10 (for stays) and 8 for boiler shells.

Henry Greenly in his book “Model Engineering” says

“In calculating the working strength of a cylindrical boiler a certain factor of safety is allowed. The usual factor is 8, which means that the working pressure is only one-eighth of the pressure which the boiler would stand before
it actually failed.”

To take his “which means that the working pressure is only one-eighth of the pressure which the boiler would stand before it actually failed.” this is erroneous because the boiler will fail at the Yield Point of annealed copper, not the UTS of hard copper.

Martin Evans in “The Model Steam Locomotive” says “The factor of safety has been taken at between 6 and 10, but 8 is generally thought to be adequate”.

“thought to be adequate”?! Not proven nor demonstrated to be adequate? This is supposed to be engineering where strength of materials is a known science.

Never mind the asides, the point is that the commonly adopted ‘Safety Factors” for the UTS method are in a range of 6 to 10. Quite a wide range.

Where Did These “Safety Factors” Come From?

If anyone knows please come forward.

But I’ve discovered something quite interesting.

If the UTS of hard copper is divided by the Yield Point of annealed copper, guess what?

The resulting ratio for three sources of information yield:-

4.7, 6.4 and 10.3.

Say a range from 5 to 10.

Isn’t it interesting that these are so close to the so-called “Safety Factors” of the promoters of the UTS method of 6 to 10.

Food for thought?

More coming in the next post.

PS

Here are the sources of information and the calculations for the ratios mentioned above:-

[JasonBModerator @jasonb]

“thought to be adequate”?! Not proven nor demonstrated to be adequate? This is supposed to be engineering where strength of materials is a known science.

 

Do you not consider the lack of boiler failures built to Evans, etc, formula as proving that they are adequate?

It can be seen that your methiod gives a constant 3.5% thicker result so Evans’ formula is not that far off and if the builder decided on a slightly different SF than 8 could well end up with a thicker minimum material than your method gives so would certainly be adequate in those cases.

Given that the wall thickness tolerance of a piece of 16g tube can be +/- 0.006″ (approx 10%) in the sort of sizes lilely to be use dfor model boilers. That kind of puts the 3.5% difference of the two methods which at our sort of sizes is 1-3 thou into perspective. See towards bottom of page

 

[Paul Kemp]

YPM is not a boiler design method, it is one tool to determine one of the criteria (shell thickness)

On Paul Kemp Said:

Michael,

I do agree with Martin though that the YPM is not a boiler design method, it is one tool to determine one of the criteria (shell thickness).  There is a lot more to designing a boiler than determining the hoop stress and choosing a shell thickness.

Paul.

Paul, you’re absolutely spot on.

Couldn’t have said it better myself.

But the UTS method is totally flawed and outdated. (My opinion)

Please see:-

https://www.model-engineer.co.uk/forums/topic/boiler-design-issue-4765/page/6/#post-801540

 

 

[JasonB]

On JasonB Said:

“thought to be adequate”?! Not proven nor demonstrated to be adequate? This is supposed to be engineering where strength of materials is a known science.

 

Do you not consider the lack of boiler failures built to Evans, etc, formula as proving that they are adequate?

It can be seen that your methiod gives a constant 3.5% thicker result so Evans’ formula is not that far off and if the builder decided on a slightly different SF than 8 could well end up with a thicker minimum material than your method gives so would certainly be adequate in those cases.

Given that the wall thickness tolerance of a piece of 16g tube can be +/- 0.006″ (approx 10%) in the sort of sizes lilely to be use dfor model boilers. That kind of puts the 3.5% difference of the two methods which at our sort of sizes is 1-3 thou into perspective. See towards bottom of page

 

JasonB as stated many times lack of reported failures does not prove that the design method is correct. The UTS method is fatally lawed see The UTS Method Explained – Part 2 – Fudge Factors

[MEinThailand]

On MEinThailand Said:

On JasonB Said:

“thought to be adequate”?! Not proven nor demonstrated to be adequate? This is supposed to be engineering where strength of materials is a known science.

 

Do you not consider the lack of boiler failures built to Evans, etc, formula as proving that they are adequate?

It can be seen that your methiod gives a constant 3.5% thicker result so Evans’ formula is not that far off and if the builder decided on a slightly different SF than 8 could well end up with a thicker minimum material than your method gives so would certainly be adequate in those cases.

Given that the wall thickness tolerance of a piece of 16g tube can be +/- 0.006″ (approx 10%) in the sort of sizes lilely to be use dfor model boilers. That kind of puts the 3.5% difference of the two methods which at our sort of sizes is 1-3 thou into perspective. See towards bottom of page

 

JasonB as stated many times lack of reported failures does not prove that the design method is correct. The UTS method is fatally lawed see The UTS Method Explained – Part 2 – Fudge Factors

Hi JasonB I think I owe you an apology. It seems you were referring to my post I referenced above. Unfortunately I don’t have all day just to sit here and review/criticise other people’s posts. I’ve got other things to do.

By the way, the Yield Point Method, is not my method. It is an engineering method.

And “thought to be adequate” just opens another can of worms. “thought to be adequate” by whom? Where is the engineering data to support that?

There is a difference between the results of a calculation and the method used. Two different methods may give the same answer numerically, but that does not prove that the two different methods used are correct. The UTS method is incorrect from both a logical and an engineering perspective.

[derek hall 1Participant @derekhall1]

I thought I read somewhere, or maybe a fell asleep in a lecture during my engineering training, that some materials change their properties as they get older. I am sure aluminium does.

Doesn’t copper also “age harden”?, if so how does that factor into this discussion?, or is this “change” too trivial?

 

I would love to know what T D Walshaw (Tubal Cain) would have bought to this discussion 😊

[Clive Brown 1Participant @clivebrown1]

Age hardening is a two-stage heat treatment process applied to suitable alloys, including some of aluminium and copper. It doesn’t occur in metals in their pure form and so is not relevant to normal model boiler-making.

It doesn’t refer to materials just getting older. AFAIK, the only way to harden pure copper is to mechanically deform it.

[Clive Brown 1]

On Clive Brown 1 Said:

Age hardening is a two-stage heat treatment process applied to suitable alloys, including some of aluminium and copper. It doesn’t occur in metals in their pure form and so is not relevant to normal model boiler-making.

It doesn’t refer to materials just getting older. AFAIK, the only way to harden pure copper is to mechanically deform it.

Hi Clive Brown 1,

I was thinking along the same lines as you that “the only way to harden pure copper is to mechanically deform it.”

But I recall a post from Luker who said:-

“During the test the material deforms and the stresses are relaxed, the material also goes through isotropic hardening, resulting in a stronger boiler. I actually simulated one of my boilers using a non-linear hardening material model and the strength improvement is significant (this was presented at an international FEM conference, due to the complexities of the coupled physics and linked model states).”

I don’t understand what Luker was describing, but anyway it’s outside the scope of our article which only goes as far as as describing a way to calculate the minimum thickness of coper boiler shells.

Perhaps Luker has something to enlighten us with?

[JasonBModerator @jasonb]

As has been said here a few times the UTS method is a simplified form of calculation and the linear-based calculations aren’t recognised in industry. As LukeR is in the industry, I expect he understands these things better than most of us.

As most of the boiler committees around the world comprise of people who have direct experience in the pressure vessle industry or those who have made a lot of boilers themselves, it is them you will have to convince if the UTS method is to be adopted by the committees and boiler inspectors.

The same goes with the poeple who’s formulas we have been using for years. Being in the design office of companies like Cammell Laird and Joseph Anderson Boiler Makers would tend to have given them a good understanding of the design of boilers in industry and applied that to our models.

[JasonBModerator @jasonb]

Interesting to watch one of LukeR’s older boiler simulations, the actual barrel hardly changes colour so is really not the main part of the boiler to be thinking about if you want to come up with a safer option. Which again brings up the question is a change in method of barrel calculation really needed and even if it was what actual effect would it have?

[Michael GilliganParticipant @michaelgilligan61133]

Having already made it clear that I have no real personal interest in boilers …

As an onlooker, I am finding this discussion most enlightening.

There does, however, seem to be a significant lack of empirical evidence: So might I suggest that it’s time for someone with the appropriate facilities to conduct some destructive tests ?

We need to see a few samples of ‘real world’ copper tube, hydraulically tested to failure.

MichaelG.

[noel shelleyParticipant @noelshelley55608]

The rolled plate of my 1944 air compressor tank is still good for a 1.5 xWP test and due to a problem stood over 350psi at one point. The 11 year old tank on a modern compressor failed due to corrosion, the clue was condensation puddles under the tank.  The lighter and much thinner tank stood it’s tests – but not for long ! Modern methods ?

By calculation modern items are designed to pass inspection, be easier to use, lighter, faster, better ? Break and be thrown away !

A lack of reported failures of copper model boilers seems to indicate that the designs and how they were built were sound ?

Not proven nor demonstrated to be adequate – how many decades of proof is needed ?

That the authors are passionate about the matter, if they want change then they need to come up with REAL boilers that clearly show the benefits and the need for change. I view the case as “not proven” – yet.  Noel.

[JasonBModerator @jasonb]

Michael I did ask for similar in my 1st or 2nd reply. Even happy to accept FEA if cost or other factors prevent real copper boilers being made.

There is an article in ME4651 that shows some test boilers pushed to bursting, the failure is mostly around bushes. It would have been interesting to see some samples with the inlet on one of the endplates so nothing to affect the tube. Photo from article

If I get a moment I might run the two calculation methods to see what the minimum thickness for thos eexamples works out at.

[Clive Brown 1Participant @clivebrown1]

JasonB really beat me to it with his convincing examples of relatively thin shells subjected to high test pressures. i was going to say that the typical model boiler, as published by ME over many years, is of very conservative design with quite low working stresses. Eg; for my Allchin, hoop stress in the shell at 100psi pressure is around 2000psi which is almost negligible. Unless badly mishandled, such a model boiler therefore, if it were fail, would probably fail at a badly made joint, not by failure of the material. This justifies the practice of employing club boiler inspectors who, while not themselves expert in pressure vessel design, can assess the quality of workmanship and adherance to specification to an extent acceptable to the insurance industry.

[JasonB]

On JasonB Said:

Michael I did ask for similar in my 1st or 2nd reply. […]

I accept that, Jason … but it’s not quite what I was suggesting

Those were boilers, but I think it would be worth backing-off a little, and simply collecting more evidence … from many samples of simple tubes of various sizes

[it’s mathematically simple to normalise the results]

MichaelG.

[MEinThailand]

On MEinThailand Said:

The UTS Method Explained – Part 1 – Close Light

A Very Funny (But True) Story About The Thai Language And How It Relates To The UTS Method

…So pit fai means close light. Shut the light off. They never say switch the light off. Just close light.

In my big workshop I have a 3 phase distribution board with (of course) an isolator or as the Thai say a ‘cut-out’. It isolated the DB and hence the workshop. I always switch off all the lights and isolate the DB before locking up at night and my Thai wife always asks me if I’ve done that in case I might have forgotten.

Two days ago after I had locked up for the night she asked the usual question “Pit cut-out mai?”. Mai just indicates a question and as we’ve seen above, Pit means to close, hence; “Close cut-out mai?” intending to mean did I switch of the power at the cut off? But literally it means “Close cut-out?”

Feeling a bit mischievous I said “No, I opened the cut-out”.

Of course she got angry and demanded that I unlock the workshop and “Pit cut-out” – close the cut-out.

As I tried to explain that the Thais have it all wrong in relation to switches and electricity she got even more agitated and angry. I tried to explain that you have open an (electrical) circuit at the switch or cut-out in order to stop the electricity flowing. In order to “pit fai” (close light) you have to pert (open) (not pit – close) the cut-out.

I refused to reopen the workshop and “close cut-out” and instead tried to explain that in order to switch off the power you have to OPEN the cut-out. She wouldn’t have it. And a flaming row erupted. She insisted that I “close cut-out”.

What has this got to do with the UTS Method for boiler design?

It just shows how hard it is get someone to change their way at looking for something when they’ve always done (or said) it that way for years. When it’s become engrained over many years it’s very difficult to show people a new way of looking at things. It’s just human nature.

…Thank you.

Now I’m worried.  MEinThailand mischievously started a flaming row with the wife over a linguistic difference! On the web that behaviour is called trolling.  A bad example too, because it’s not that English is right and Thai is wrong.   Nor is wifey foolishly stubborn!   In her experience, it’s hubby who is upside down.

George Boole showed the equivalence between positive and negative logic.  That is, TRUE/FALSE, HIGH/LOW, ON/OFF and OPEN/CLOSED can be expressed either way round.   Curiously when designing digital circuits, it’s often simpler to work in NAND, NOR negative logic than the more intuitive AND, OR.  For that reason many digital ICs are negative logic.

Boole published “The Laws of Thought” in 1853, and his preface contains this relevant observation:

He was fully aware that learned and able men maintained opinions
upon the subject of Logic directly opposed to the views upon which the entire
argument and procedure of his work rested. While he believed those opinions to
be erroneous, he was conscious that his own views might insensibly be warped
by an influence of another kind. He felt in an especial manner the danger of that
intellectual bias which long attention to a particular aspect of truth tends to
produce. But he trusts that out of this conflict of opinions the same truth will
but emerge the more free from any personal admixture; that its different parts
will be seen in their just proportion; and that none of them will eventually be
too highly valued or too lightly regarded because of the prejudices which may
attach to the mere form of its exposition.

I hope the defence of Yield Point versus UTS isn’t built on shaky foundations!

🙂

Dave

[Clive Brown 1Participant @clivebrown1]

Michael, I don’t think testing tube samples would yield, (sorry!) much useful new information. Tensile properties of materials can be readily measured by standard tests. Formulae for pressure induced stresses in shells are well established. Standardisation of the test specimens could be awkward with results affected by end effects, material variation etc. OK, interesting but not scientifically valuable.

[Michael GilliganParticipant @michaelgilligan61133]

Clive

I have a nagging suspicion that modern manufacturing methods  might be changing the characteristics of what is sold as tube [and probably not for the better] … I fully accept that adequate tensile testing of pure copper samples must have already been done: it’s text-book stuff after all.

I will stop interfering and let all you experts fight it out.

MichaelG.

.

Ref. [one of many]
https://www.researchgate.net/publication/236629224_Analysis_of_Copper_Pitting_Failure_in_Drinking_Water_Distribution_System

[MEinThailand]

On MEinThailand Said:

…Where Did These “Safety Factors” Come From?

If anyone knows please come forward.

…

Happy to oblige!  There is nothing ‘so called’ about Safety Factors, they’re well established, dating back well over a century.

Here’s a couple of pages from JEFFERSON T.B and BROOKING W.J Introduction to Mechanical Design. New York: Ronald Press Company,   1951.  (Chapter 2)

is used only to indicate the margin between the actual stress computed as accurately as possible and the stress that will cause functional failure under operational conditions.

Dave

 

 

 

[Author]

Posts