Copper Boiler Design

This forum is dedicated to the Live Steam Hobbyist Community.

Moderators: cbrew, Harold_V

James Powell
Posts: 503
Joined: Fri Oct 28, 2011 10:42 pm

Re: Copper Boiler Design

Post by James Powell »

Caribou 1 (that I have) has superheaters that definetly do work. Proof? Simples, if you run TK 460 oil (600 weight, saturated oil), it will sieze up solid if you don't run with a high water level. If you use 1200 weight oil, then it doesn't sieze up...or, if you keep the water level right up near the top, and get enough carry over to cool the steam.

It's still no-where near as much superheat as the Doble Flash plant produced, that would give a dull glow to the superheater before we shortened it by (5-7 ft?). Unfortunately, the block on the S-50 no longer shows the beautiful colour change which it used to, as it has oxidized across the bottom to a more uniform aged gunmetal colour.

Superheat can and does produce a great difference in performance. There is also the density difference that has to be taken into account when dealing with superheated vs saturated steam. (superheated steam flows quite a bit easier than steam with a quality of less than 1.0, or fully dry).

If you have calculated the area of the tubes to be about the maximum in the space available, then that's it, and there is no point arguing the results. I _might_ consider going to 2 superheater flues & 4 more tubes, but that is kind of iffy.

As regards the arrangement of the pettycoat pipe, JJS Koofman (? sp) and Micheal Guy are probably getting on for the experts. I know that Micheal is a Toronto area indivdual.

LBSC made many designs which will work. That doesn't mean that they are the most efficient engine designs possible, just that they will work, and pull a useful load. Jim Ewins is one of the better known engineers who have worked on the problems of efficient operation of model (mostly 5" gauge) locos. There's currently a contemporary loco for sale on Station Road Steam to Jim's heavily modified 9F.
I'd be interested to play, but time is always a problem. (we've wasted 15 years between dad and I...unfortunately, my work tends to be a bit more than full time, as operating a full sized plant that has taken a lot of abuse takes a lot of effort). Results are important, and changing variables costs time in this sport.

James
User avatar
baggo
Posts: 161
Joined: Mon Jan 12, 2009 6:25 am
Location: Derby, UK

Re: Copper Boiler Design

Post by baggo »

Carrdo,

The Gwen Elms is a wide firebox design (hence the poor area ratio) with a grate area of 18 sq. inches whilst the King is narrow with a grate area of only 7 sq. inches. Both have 4 cylinders of 0.6875" bore and 1.125 inch stroke but the King is a better steamer! I think so because of the better designed boiler. The Gwen Elms needs a strong blast to drag enough air through the large grate area and the blast arrangements probably need modifying to achieve this.

I've studied Jim Ewins work in great detail and tried to simplify his formulae into easy to use spreadsheets but as has been stated before, his work was based on 5 inch gauge locos and does not always fit in with other gauges. His formuae certainly fall down on many 2.5 inch gauge designs and I believe this to be down to the tubes - you can only go so small with the tube diameter before they become impractical.

I've come to the conclusion that there are two critical areas in the boiler design:

1) The grate area should match the steam consumption of the cylinders.

2) The gas area of the tubes should match the area of the grate if possible.

Wide firebox locos tend to have grate areas that are much too large for the size of the cylinders. This makes them very easy to fire and drive but makes them very inefficient ( not really critical unless we are out to win efficiency competitions!). As James has mentioned, you can get away with a lazy fire because you have such a mass of burning coal to supply the required heat.

Narrow firebox locos often have insufficient grate area for the cylinder size and so the fires have to be run very hot to get enough heat, resulting in clinkering and burnt grates!

I am reaching the conclusion that in 'our' sizes, the L/D2 ratio doesn't really matter that much. It is generally accepted that only the first few inches of the tubes actually contribute to steam production (the majority of steam production comes from the firebox) so the remainder of the tube is just to carry the waste gases to the smokebox and contribute nothing to steam production.

The exception is when turbulators are fitted. Members of the forum have fitted them to propane and oil fired locos and noted a great improvement in steam production due to the turbulent flow they cause through the tubes increasing the heat transfer. Unfortunately, they are not really practical in a coal burner.

Keiller's ratio of 50 - 70 was mainly based on full sized locos where the conditions in the tubes are completely different to those in a model so perhaps does not apply to a model in real life.

I'm a big fan of superheat - the more the better! I think it makes a big improvement in how a loco performs besides decreasing coal and water consumption.This can cause problems with lubrication etc. as has been mentioned by James. My Helen Long has radiant superheaters that extend the full length of the firebox and the steam is HOT! I had no end of problems when I first ran her with the piston valves seizing and scoring. I eventually fitted valves with PTFE heads and have had no further problems. I did try and measure the temperature of the cylinders when the loco was running on a rolling road but my thermometer only went up to 160° C and it was way past that! I originally used O ring seals on the steam manifold joints but they just disintegrated with the heat. The steam manifold itself turned blue after the first run!

One of the locos in production at the moment is a much modified Simplex with the boiler completely redesigned to fit the JIm Ewins formulae but I won't know the results for some time yet unfortunately. It will be interesting to see how this performs against a standard Simplex which according to Jim's theories has a grate area on the small size and a tube area to grate area of only 9%. This will hopefully have a Lempor exhaust designed from Michael Guy's pages

It would be nice to be able to spend the time (and the money) trying out different boilers on the same chassis to compare results but time and money are in too short a supply!

As an aside, I think it was Keiller who said "the problem with steam locos is that no matter how badly designed or badly made, they will still work" (or words to that effect!)

John
Secretary of The National 2½" Gauge Association
Member of North West Liecestershire SME

http://www.modeng.johnbaguley.info
User avatar
Fred_V
Posts: 4370
Joined: Sun Jan 12, 2003 3:26 pm

Re: Copper Boiler Design

Post by Fred_V »

Carrdo wrote:Alan,

Your reply prompted me to look through the papers here and I did find a couple of articles by Jim Ewins namely, Model Locomotive Research A brief survey and An Experimental Model Based on the BR 2-10-0 Class of Locomotive a series which was published in ?

I will have to review the Ewins work again.

Thumbing through the series I noticed that the only wide firebox American locomotive Martin Evans ever did a construction series on, Columbia a 4-8-4 Northern in 3/4" scale has (based on Kieller) a "Tube Factor" of 98.

Coincidence?

There is much much more of course to Ewins and I remember him coming to the "conclusion" through all of his research that everything done to date in every aspect of model locomotive design and practice left an awful lot of room for improvement even with the work of such notables as LBSC (one has to read this between the lines, Ewins is too scholarly and subtle to say this directly but the results of his research certainly do).
I did some study of Ewin's work and made up a spread sheet for his formulas that i can send anyone that wants it. i build in 2.5" scale or larger and found that his numbers didn't relate at all in larger scales.
Fred V
Pensacola, Fl.
RET
Posts: 960
Joined: Wed Jun 07, 2006 8:36 am
Location: Toronto, Canada

Re: Copper Boiler Design

Post by RET »

Hi,

Its interesting to see where this discussion is going. I see several "conclusions" emerging. It is also useful to have all of this information together in one thread.

First, if "L" over "D" squared is within the "acceptable" range, it makes for a locomotive that "breathes" easier, but if it isn't, this can be easily compensated for by changing the draft in the smokebox.

Second, it seems (within reason) that the more superheat (real) the better. Remember, true superheat doesn't occur until all the water carryover and condensation have been evaporated. Apparently there are lubrication problems that have to be addressed, but they are solvable. As James says, dry, superheated steam flows through passages easier than saturated steam. This is desirable in our models and because there are no water droplets to serve as condensation nucleii, the steam could go a little below the condensation point before droplets actually appeared.

To get this level of superheat, "radiant" superheaters which extend into the firebox must be used, the "Curly" type that are only in the flues as used in full size just can't do the job. These superheaters must be welded stainless steel to stand up to the firebox temperatures, especially when coal is used.

There may be more, but this is what I see now.

Richard Trounce.
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Copper Boiler Design

Post by Carrdo »

I don't want to keep up the numbers game but I went through the Ewins articles and I have and came up with the following values for the Atlantic boiler and engine as they both exist now compared to Ewins suggested values taken from well proven designs .

To give Ewins credit, he does say and repeats it often in his articles that his numbers and data is taken only from 3-1/2" and 5" gauge model locomotives and that to extend his data to other gauges may not be meaningful. A great deal of of additional data needs to be collected in these other gauges first to have meaningful comparisons.

I also want to draw my own very simplified conclusions (which you can comment on or criticize) as to what this all means for my boiler and engine combination as Ewins rightly says the two have to work together with boiler steam output matching engine demand and consumption under most or all of the operating, track and driver combinations.

Easy to say but not easy to do.

But before one can calculate the engine factor (Ee), one has to have the basic engine dimensions. They are:

Cylinder Bore -1-5/8"
Cylinder stroke -2-1/2"
Driving wheel diameter- 6-5/8"

Based on the above and the boiler data outlined previously, here are the numbers.

Boiler tube Factor (Kt) - 99, suggested 65-70 by Kieller, 80 noted by Ewins

Boiler factor (Eb) - 238, 80 suggested by Ewins

Engine Factor (Ee) 0.065, 0.12-0.25 suggested by Ewins

Overall Factor (Eo) 15.47, 10-14 suggested by Ewins

with the proviso that the factor Eo is meaningful only if Eb and Ee are also close to his suggested values which it turns out they aren't in our case.

So what does all this mean?

To me, in my simple way of thinking, the Atlantic is "overboilered" that is with the wide firebox and huge grate area, the boiler can produce much more steam than the engine can use except if it is worked hard that is, it is pulling a very heavy load and is travelling at high speed.

The corollarly to this is there will be a lazy fire in the firebox most of the time as Jamie Powell points out and there may be great difficulty in keeping the fire going at all under light loads and slow speeds. To add to this, the present tube size may be restricting the draft. This is also bad news, as a lazy fire does need added draft to keep it going so one may have to run with the blower on.

Unfortunately, for the most part, we can't pull truly heavy loads as models in this size cannot develope anywhere near their potential tractive effort as the drivers will slip long before this is achieved due to the scale effect (i.e. weight in general and weight on the drivers in particular decreases much faster than size).

Even Ewins 9F, which is a big locomotive in 5" gauge, could only achieve a tractive effort of about 0.3 (before massive wheel slip set in) when he deliberately tried to steam the 9F boiler to its limit (he never did find the boilers true limit because of this)(and this was after he made his specially constructed heavy box compartment main frames where the box compartments were filled with lead so that his locomotive ended up having about twice the weight on the drivers as a normally constructed model).

Full size steam was designed on a ratio of weight on drivers/maximum tractive effort of over 4, at least over here for the "Superpower" locomotives.

To add to all of this, because of the way I drive, the Atlantic will definitely be "loafing around" all of the time. I guess I will have to learn to be a more assertive/aggressive driver.

The Atlantic is what Ewins classifies as a large express passenger type of locomotive with large firebox and large wheels. It will deviate greatly from his suggested values and he predicts it will have all of the problems outlined above as it is an "outlier".

But as Richard says we can get it to work and work well by dong things which don't have to be considered in the larger gauges or in full size practice such as the use of the double petticoat to induce better drafting conditions in the smokebox.

If I were making this boiler from scratch again as a saturated steamer only, I would, as I have said before, consider going up a tube size to nominal 1/2" OD copper tubing with as many as I could but likely fewer tubes.

I like Bill's practice of calculating tube size on dirty tubes after 20 minutes of steaming but I am counting on Richard smokebox improvements to keep the tubes pretty clean. I think they will be: with the way I drive I can't see anything I will do that will lift the fire!

Anyway, all of this is heading towards "paralysis by analysis" so first get the boiler job done and then see what happens.

As a postscript, I think Ewins research makes for fascinating reading even if you don't agree with what he is saying or see the results he predicted. He has some very different and indeed many would say radical ideas which extend to every aspect of model live steam locomotive design.

New ideas and methods are always controversial but he is well worth a read despite this. It will open ones eyes to other possibilities and other approaches there are if nothing else.
Last edited by Carrdo on Tue Dec 11, 2012 10:43 am, edited 6 times in total.
User avatar
LVRR2095
Posts: 1688
Joined: Sun Mar 04, 2007 6:50 pm
Location: Maine, USA

Re: Copper Boiler Design

Post by LVRR2095 »

Carrdo wrote: If I were making this boiler from scratch again as a saturated steamer only, I would, as I have said before, consider going up a tube size to nominal 1/2" OD copper tubing with as many as I could but likely fewer tubes.

.
Don....just as a bit of trivia....the 2 - 1/2" gauge Coventry B&O P-7 Pacific uses seven 1/2" o.d. fire tubes.
I can tell you that even a non-superheated boiler... 3 - 1/2" o.d. boiler barrel, those P-7's steam like witches and will run faster than I care to go!
Keith
Asteamhead
Posts: 582
Joined: Tue Jul 31, 2012 1:59 pm
Location: Germany, Duesseldorf

Re: Copper Boiler Design

Post by Asteamhead »

baggo wrote:Carrdo,

I've come to the conclusion that there are two critical areas in the boiler design:

1) The grate area should match the steam consumption of the cylinders.

2) The gas area of the tubes should match the area of the grate if possible.

I am reaching the conclusion that in 'our' sizes, the L/D2 ratio doesn't really matter that much. It is generally accepted that only the first few inches of the tubes actually contribute to steam production (the majority of steam production comes from the firebox) so the remainder of the tube is just to carry the waste gases to the smokebox and contribute nothing to steam production.

The exception is when turbulators are fitted. Members of the forum have fitted them to propane and oil fired locos and noted a great improvement in steam production due to the turbulent flow they cause through the tubes increasing the heat transfer. Unfortunately, they are not really practical in a coal burner.

I'm a big fan of superheat - the more the better! I think it makes a big improvement in how a loco performs besides decreasing coal and water consumption.This can cause problems with lubrication etc. as has been mentioned by James. My Helen Long has radiant superheaters that extend the full length of the firebox and the steam is HOT! I had no end of problems when I first ran her with the piston valves seizing and scoring. I eventually fitted valves with PTFE heads and have had no further problems. I did try and measure the temperature of the cylinders when the loco was running on a rolling road but my thermometer only went up to 160° C and it was way past that! I originally used O ring seals on the steam manifold joints but they just disintegrated with the heat. The steam manifold itself turned blue after the first run!

It would be nice to be able to spend the time (and the money) trying out different boilers on the same chassis to compare results but time and money are in too short a supply!

As an aside, I think it was Keiller who said "the problem with steam locos is that no matter how badly designed or badly made, they will still work" (or words to that effect!)

John
Hello John,
to "fire on" this discussion here are some comments to your statements:
You are absolutely right regarding your design rules (size of grate and gas area)!
But I don't agree to your conclusion regarding heat transfers in the firebox versus tubes and flues! This is told on and on, but a rough calculation based on physics shows a different view! Please see my attached paper (sorry just written by hand). Hope you may follow my thoughts.
Evaporation of the firebox is about 50 % of total - otherwise effiency of a locomotive boiler would be very ppor! The bigger the boiler, the more % of total heat is transferred by the tubes / flues. Look, the superheater alone will carry up to 1 / 5 of total heat @ 400 °C superheated steam.
This is leading to my last point regarding proper superheating. As long as the factor of free gas area in both tubes and flues incl. superheater devided by total heating surface in each of them is about the same, you will achieve ample superheating. No superheater in the firebox necessary! Please don't forget the temperature drop from the outside of the superheater tube and the steam inside. Superheated steam is a poor heat conductor thus velocity of steam is to be very high to get a sufficient heat transfer. Many superheaters are constructed with tubes too big in diameter (don't be too anxious of preasure drop inside!). In addition the free gas area through the flues might be throttled too much.
When I started building my DB class 44 in 1 in 10 scale back in 1977, it had a superheater in 8 flues each with 2 tubes. Believe me, the steam temperature rose to about unwanted 320 °C! No viton O-ring nor any teflon could be used reliable for the steam passages. Thus I decreased length of superheater tubes slightly to come down to not more than 300 °C at full load. But one good news: Superheated steam works fine for pumps and blower too!

And your last sentence tales the story in short - similar to as I express it: "To run a steam engine all you need is steam, steam and steam again and everything else doesn't matter!"

Asteamhead
Attachments
Boiler calc 1.pdf
(311.42 KiB) Downloaded 266 times
Boiler calc 2.pdf
(256.09 KiB) Downloaded 245 times
Boiler calc 3.pdf
(310.6 KiB) Downloaded 224 times
User avatar
tsph6500
Posts: 1417
Joined: Fri Dec 03, 2004 3:38 pm
Location: West of Dunvegan, Ontario CANADA
Contact:

Re: Copper Boiler Design

Post by tsph6500 »

Maybe I've missed it in this thread so far, numbers and formulae make me swoon… but I found the Australian Bolier Code to have some good practices. Like all of the volumes on boilermaking mentioned, none are perfect but together they give one a good guide to the subject. So does looking at what works on the next steaming bay.
Best regards,
Jim Leggett

Montreal Live Steamers
www.montreallivesteamers.org

A Founding Member of the Tinkerbell Scale Society - Northern Division
I'm an A.R.S.E. (Association of Railroad Steam Engineers)
Toad Swamp & Punk Hollow Railroad - Head Tycoon
The Juvenile Traction Company - CEO & Apprentice Machinist 3rd Class
White Mountain Central RR - Engineer & Fireman
James Powell
Posts: 503
Joined: Fri Oct 28, 2011 10:42 pm

Re: Copper Boiler Design

Post by James Powell »

Jim,

You've seen the difference between "adaquate" and "_wow_" performance- namely, the difference between Caribou 1 and Robin Hood. Is it worth the effort? Questionable, at best, in a 3.5" gauge model. Even in a 7.x gauge model, it is unlikely to matter, as it isn't being used to earn $$$ for the owner, and even if it was, then the cost of fuel is a very minor portion of the overall cost of ownership. (I've never heard of a live steam engine owner complain about the cost of fuel in comparison to the # of hours they spent building the model...certainly, I would rather spend $ than time on models).

I want a model which will run reliably first, and then efficiently a distant 2nd. 1 is important, 2 is rather less so, especially if I am burning coal. How long do you think I could steam with the 4.5 tons of coal I have outside with Caribou ? (hmm...3 lb/hr, that gives 140 days of continous operation...I think the wheels would need turning at some point !). That is ~$1800 worth of coal, so...what, figure something like 10 dollars per 24 hours, or $.50/hr of operating for coal...the fuel for the engineer would be more expensive. (timbits are now .20/each, figure 3/hr, or .60/hr minimum for the engineer...and coffee...better get an IV supply and a drive through window (motorized vehicle !) for me...)

So, basically, N is less important than E, by at least 50%. I'd be wary of tubes smaller than 7/16, and probably go with 1/2 out of preference. (& that's in 3.5" gauge). The L/D ratio seems more critical than the free gas area, but both probably have an optimum value for any given coal.

James
User avatar
Fred_V
Posts: 4370
Joined: Sun Jan 12, 2003 3:26 pm

Re: Copper Boiler Design

Post by Fred_V »

Asteamhead wrote:
Hello John,
to "fire on" this discussion here are some comments to your statements:
You are absolutely right regarding your design rules (size of grate and gas area)!
But I don't agree to your conclusion regarding heat transfers in the firebox versus tubes and flues! This is told on and on, but a rough calculation based on physics shows a different view! Please see my attached paper (sorry just written by hand). Hope you may follow my thoughts.

Asteamhead
I have always questioned that statement too. look at vertical boilers. also a guy in south Fl. built a std looking boiler but it had no firebox. he had 2 large flues with burners in them. it was a shy steamer so he added a third flue and it ran fine.
Fred V
Pensacola, Fl.
alanstepney
Posts: 652
Joined: Tue Oct 28, 2003 10:56 am

Re: Copper Boiler Design

Post by alanstepney »

I am pleased to see some attempts by various people to give some meaningful numbers to model loco boilers.

Jim Ewins work was excellent and gave us a decent basis to work from, but with limitations (As Carddo said).

Gathering the information required to cater for other gauges & scales, would be interesting and may well prove fruitful, but would be a massive task.

The other comments and figures given in this thread are also useful.

Whilst I have the greatest admiration for LBSC and other early designers, I also like to see a more scientific approach, which can only benefit all of us in this hobby.
http://www.alanstepney.info
Model Engineering, Steam and workshop pages.
David Powell
Posts: 523
Joined: Thu Mar 13, 2003 10:38 pm
Location: Pickering Ontario Canada.

Re: Copper Boiler Design

Post by David Powell »

We once followed a " scientific " approach. I ended up building a boiler for Mhari's steam roller with an approximately 8 sq in firegrate and relatively long 3/8 tubes. We ran it for about 20 yrs, both coal and sometimes propane fired. It was a misery, a poor steamer, with little ability to recover if steam pressure dropped when coal fired, seemed to be incapable of giving" that bit extra" when needed, and the tubes needed frequent cleaning. The replacement has 12 sq " of grate, 1/2 od tubes and a smokebox coil dryer. It is easy to steam. will burn almost any coal,the harder you drive the more steam it will make, it will pull up from 20 lbs to blowing off while idling round or pulling the driver on the trailer and does not need tube cleaning every couple of hours of steaming. Both are steel boilers with copper tubes. Perhaps the second boiler is less" efficient", the chimney seems hotter now, BUT at least it is " effective" in doing its job. Regards David Powell.
Post Reply