99 3462 Build diary

[mbrown]

I have decided to try and resolve the problem myself... so I have shelled out on a cyclone burner for propane. Unfortunately, you can't get them to go with my Bullfinch equipment, so it has meant getting a Sievert handle as well. I gather a cyclone burner will operate in a confined space like a firebox.

I am very conscious that authorities like Alec Farmer advise against trying to do further silver soldering on a completed boiler. But as I don't want any soft solder in the firebox, the boiler is either a write-off or I will have to find a way to re-solder it with Easyflo. Never having used oxy-acetylene, I don't want to attempt using it on this, and I am not keen to let someone else have a go if they haven't experience of model boilers.

The way I intend to approach it is to use my big Bullfinch torch to heat the whole boiler as evenly as possible to something below the melting point of the solder, and use the cyclone burner on a second torch to give local heat to the corner of the firebox. Then let it all cool very slowly.... with luck, I may avoid differential expansion and undue stresses.

I will go on pondering as it will be a few days before the cylcone burner arrives. In the meantime, any advice from the forum (even "Don't even attempt this!") will be welcome.

Malcolm

[don9f]

Hi Malcolm, I’m no expert but I don’t see any reason why your proposal shouldn’t work.
Many years ago I helped a fellow modeller with a 5” Black Five boiler that had a leaking superheater flue on the firebox tubeplate (found like yours, after initial test). With background heat and a gentle oxy acetylene flame, the defective joint was re-soldered, the boiler cooled very slowly and it never gave any more trouble.

Cheers Don

[oldnorton]

Hi Malcolm

I am just joining in the conversation to give a couple of thoughts. The big cyclone burners are on long stems to keep your hand away from the heat, but the air is still drawn for the burner head behind the gas jet (well it is on the two I have). That means it will still extinguish in an enclosed space. That is why oxy-propane or acetylene works with its own oxygen supply. It might be worth getting a small bottle kit, but you will need some rehearsal. Put a (competent) friend on the big propane keeping parts background hot then you on the oxy-propane going in on the problem area. Just as Don's description I did the same on a big 5" boiler. Had to chase a few follow-on leaks with stays but got there in the end. We made an attempt to slowly get heat up the tubes while also heating the barrel, then blanketed the barrel to work on the firebox, then bricked it all to fully cover for slow cooling.

Norm

[flyingfox]

Greetings Malcolm, bad luck with your boiler, but keep going and I am sure things will turn out OK.
My I suggest that you go overboard in cleaning the affected area, put plenty of flux inside, so that if you have the misfortune to melt any joints, they will reform under the molten flux. Finally, I would attempt, whilst the copper is soft, to close the gap that you have as much as possible, to reduce the amount the solder has to seal/

I am not sure if you have ever seen a cyclone burner running, it gives a ball of flame, which may heat a larger area than you might want, so I would suggest that you consider oxy acetylene, or oxy propane, inside the firebox, after preheating the boiler as you describe. Slow cooling is also a good idea.

Regards
Brian B

[mbrown]

Thanks guys.

Well, I have ordered the cyclone burner so I will give it a try when it comes. According to the Cupalloys website they can be used in enclosed spaces "such as fireboxes". I have always found advice from Mr Cupalloys (can't remember his name, sorry!, but I know he is a member here) to be reliable. The Sievert cyclone burner seems to take in air from quite near the handle, so it might work...

And yes, I intend to slather flux over every joint, just in case.

It may take a week or two, but I will report on progress when there is any!

Meanwhile, there are plenty of other things to do. Today I made up a new, screw in, elbow for the top watergauge fitting and turned up a joint ring to fit the boiler and smokebox together.

Best wishes,

Malcolm

[Roger]

Dec 28, 2021 15:53:14 GMT mbrown said:

Oh dear....

Well, I plugged up all the bushes, the regulator flange and the wet header, and fitted the dome. Then I rigged up the test gear...

Result - I have a nasty leak at the top corner of the inner firebox tubeplate - about the most inaccessible spot you could imagine. This was a perfectly good joint when inspected, but what seems to have happened is that, during the staying, the solder has run out and left a small gap. There is a stay head very close to the joint.

So what to do? This is the first time a boiler of mine hasn't been tight from the word go - the worst I have had to deal with in the past was a tiny pin head leak on an outside stay. Clearly I can't re-solder it with propane now that the firebox is complete. I must see if I can find someone with oxy-acetylene who could tackle it. I must start asking around....

Hey ho. I hope it is remediable - the thought of starting the boiler all over again is too demoralising to contemplate. I think I would take to drink if that happened....

Malcolm

Hi Malcolm,
I don't think this is going to be a particularly difficult problem to address if you can enlist the assistance of someone with an Oxy-Acetylene torch and, more importantly, experience of building boilers with it. When my friend Silver Soldered by boiler, there were several awkward spots which had to be touched up, and he made it look easy.
The key thing is to get the whole thing beyond the 'flushing' colours temperature with big Propane torches all over, then just touch up those spots with the Oxy-Acetylene. It takes next to no time to just lift the temperature locally to re-melt the joint. You can do this many times over if necessary, mine took quite a few to address all of the issues.
Propane is fine until it goes wrong, and then you really have got a problem.

[ettingtonliam]

Malcolm
My heart goes out to you with this problem, though I can't offer any sage advice.

If all else fails, theres always Radweld-----.

[mbrown]

Thanks Roger and Richard,

I have had splendidly helpful support and advice through this forum - very grateful.

The view of the most experienced adviser is that oxy acetylene is the ideal approach but that my plan for using a cyclone burner ought to be OK if taken carefully. The cylcone burner hasn't arrived yet, so that option remains academic for now...

I have a Plan A, a Plan B and a Plan C at present. I intend to follow up on Plan A tomorrow and hold the other two approaches in reserve. I will report back and explain what they are when I know if any of them work!

Best wishes

Malcolm

[flyingfox]

Good Luck, Malcolm.

[mbrown]

Success!


With enormous thanks to the members of this forum who offered advice and moral support - the problem is fixed. Plan A worked!

Plan A was to persuade my friend with the oxy acetylene kit to let me use it myself - which he did very readily. I hadn't used oxy acetylene before, but I have seen it in use pretty often. So, rather heart in mouth, I went over to his place this morning. I fluxed up every joint I could see and used my big propane torch to pre heat it all. Not sure if I gave it enough heat, but the flux was glazing over and I didn't want to get too near the solder melting point. I then laid a length of Easyflo against the joint that had leaked, put a ring of solder around the adjacent stay head ("for luck" really) and got to work with the torch. It took longer than I had expected, but the solder seemed to melt nicely.

So back home, I pickled it, washed it, bunged up all the bushes etc - and put on some pressure. No good. The joint still leaked. But with the flux out of the way, it was clear that the solder had melted against the joint to look like a fillet, but it hadn't penetrated the joint at all. So, this afternoon, I went back, John kindly opened up his workshop again, and we went through the process a second time, this time using a No.7 tip instead of a No.5. That got the joint hot much more quickly and it was possible to see the solder flowing much better. Having gained a bit of confidence, I even topped up a seam I wasn't 100% happy with and re-did a couple of stay heads that also looked less than perfect.

Back home, clean it all up, put all the bungs back in etc. etc.... and connect the pump. No problem at all getting to 160 psi and to hold it with a stop cock on the line from the pump so that leaking ball valves didn't give a misleading result. After half an hour, it had dropped back to 140, but there was a tiny weep at the dome gasket joint - tightening the screws a touch reduced that, but not totally. Nevertheless, this time it took the best part of an hour to drop to 140, with a tear drop still showing at the dome gasket. I really need to put proper studs in that joint which will enable it to be tightened down better - the screws are temporary.

Anyway, here it is under test with a steady 150 psi on the gauge.


IMG_20220101_162034 by malcolm brown, on Flickr


You can imagine that this result is a great relief. And I have gained confidence in using oxy acetylene into the bargain. I may even have been in time to cancel the order for the cylcone burner which seems not to have been dispatched yet!

Thanks again to all who expressed support, offered advice and generally rallied round - much appreciated.

Malcolm

[dhamblin]

Well done Malcolm, very much a "Happy New Year" with that success.

Regards,

Dan

[don9f]

Yes, well done Malcolm....onwards and upwards !
Look forward very much to more on this project in 2022.

Cheers Don

[Roger]

Excellent news, that's so satisfying. I'm pretty sure you didn't pre-heat the boiler anywhere near enough if it took a lot of time with the Oxy-Acetylene. You need to get the boiler to the point just beyond where you see the colours flashing and changing as you heat it. You won't melt anything using a Propane torch waving over the outside like that.

Once it's that hot, just a quick run around each joint, say for two or three seconds is all it takes to melt the joint. When my friend did my boiler, as soon as it took more than four or five seconds to melt the joint, we got back on the Propane torches again. He managed to do the whole of the backhead with just one re-heat. The advantage of doing it this way is that you're keeping the differential stresses to a minimum and you're using much less of the more expensive gasses.

Another thing I forgot to mention is that the flame is best adjusted to give a fluffy soft flame rather than a fine sharp point. That spreads the heat over a larger area and reduces the risk of overheating or burning a hole!

Anyway, the result is what really matters.

[mbrown]

Many thanks Roger - that's all helpful knowledge to store up for the future!

Best wishes

Malcolm

[flyingfox]

Congratulations Malcolm, a job well done, and a fine result. I have often speculated if the slight pressure drop you noted is not due to any trapped air bubbles in the boiler when full of water slowly dissolving in the water and thus slowly reducing pressure displayed?
A great new year present for you.
regards
Brian B

[oldnorton]

Well done Malcolm, you are an oxygen-flame advocate now!

With the boiler well sealed the pressure goes up and down with temperature change.

Norm

[mbrown]

Fly Cranks

With the boiler completed, it is back to the running gear. I am going to have to re-make the spring pins as the very fine 3 mm threads I used were cut too deep and the nuts won't hold. But making another 16 simple but repetitive parts didn't appeal, so I started on the fly cranks instead.

Here are the wheel sets of the prototype during its current overhaul (courtesy of 99 3462's own Facebook page!). In the photo, the axles are in the wrong order - 2nd, 1st, 3rd and 4th reading left to right.

Incidentally, there is something distinctly odd about the 4th axle set, although one has to look very closely. I will treat it as a New Year quiz and see if anyone can spot the oddity - and even perhaps explain it, which I can't!


203123825_3258624701047385_5538263824916745246_n by malcolm brown, on Flickr


Anyway, I got the basic shapes water-jet cut some time ago, although the cutting was pretty rough and each one took about 30 mins to fettle up with files and emery. The sheet they were cut from had plenty of mill scale on it, but the phosphoric acid treatment, advocated on these pages, worked a treat to remove it. I had a centre mark for the axle holes and, having set each crank in the 4-jaw with the mark running accurately, they were drilled, bored and reamed for the 1/2" dia axle.


IMG_20211231_173707 by malcolm brown, on Flickr


Each wheel set has slightly different cranks. The first axle has quite thin cranks, the second has the same thickness but with additional thickness on the balance weights, the third (driving) set are very thick and the fourth in between! The first, second and fourth were machined to the relevant thickness and shaped plates (1 mm thick) silver soldered onto the cranks for the second axle. I got the scallop by drilling a 7/8" hole in the plate with a Conecut and trimming to size.


IMG_20220101_130117 by malcolm brown, on Flickr


The rivets appear visible but are imperceptible to the touch.

I made a simple jig to ensure the crankpin holes were all at an identical throw. They are 5/16" diameter except for the driving crank pins which are 3./8" dia.


IMG_20220101_162444 by malcolm brown, on Flickr


And here they all are....


IMG_20220101_164654 by malcolm brown, on Flickr


I won't fit the crankpins until I have designed the valve gear. The return cranks are fitted to squares on the driving crankpins and I intend to fix them to the pin then set the pins in the fly cranks with the pin for the eccentric rod at the correct throw. So to do that, I need to know the exact throw of the pin...


But adding the fly cranks helps give an impression of progress! Here's how she is looking now.


IMG_20220102_144929 by malcolm brown, on Flickr


Here endeth the model engineering activities of the Christmas and New Year break... domestic duties call tomorrow then it is back to the day job with a vengeance.

Best wishes

Malcolm

[don9f]

Jan 2, 2022 17:03:44 GMT mbrown said:

Incidentally, there is something distinctly odd about the 4th axle set, although one has to look very closely. I will treat it as a New Year quiz and see if anyone can spot the oddity - and even perhaps explain it, which I can't!


203123825_3258624701047385_5538263824916745246_n by malcolm brown, on Flickr

Best wishes

Malcolm

Hi Malcolm, is it to do with the orientation of the flycranks to the holes in the wheelcentres?

Cheers   Don

[lennart]

Very neat work. The proportions of your model look very accurate so far.

On the 4th wheelset I spotted two possible oddities:

1) two studs in the crank pins.

These are supposed to take small return cranks to reduce the crank throw for the lubricator drive. The lubricators appear to be mounted on the footplate somewere between the cab side plates and the firebox.

2) the flycranks and the balance weights on the wheels are not aligned.

This results in the effective centre of the balance weights being offset at an angle somewhere between the flycrank's and the wheel's balance weights. This makes sense to me on 3- or 4-cylinder engines where the wheels also balance the inside cylinders. Although it doesn't make sense to me, other 2-cylinder engines also had offset balance weights like 99 3462. At least the Thompson B1 and the german DR-class 23.10 come to my mind. I've got a (uncomplete) set of works drawings for the latter and spotted said oddity some time ago. The drawing for the driving wheel set can be seen here:

www.zander-heba.de/epages/17786094.sf/de_DE/?ObjectPath=/Shops/17786094/Products/11840

I couldn't see how the balance weights on 99 3462's 4th wheel set are set exactly but on said class 23.10 the right one is advanced and the left one is delayed by something like 6 or 7°. Therefore the "quatering" angle of the balanmce weights ends up a good bit larger than 90°.

As I've said, I have seen this a couple of times but can't explain it.

[mbrown]

Yes, you both spotted it. On the first and second axles, the balance weights on the wheels and the flycranks align nicely. On the driving axle they look a little bit out, although I can't be certain. But on the fourth axle they are way out.

I agree that some standard gauge locos have balance weights that are not exactly opposite the crank pins, but this is to do with complex balancing dynamics. As 99 3462 has balance weights cast into the wheels and also on the fly cranks, I don't see the point of having them out of phase on the same axle, especially when the other axles have the weights aligned. On the MPSB, the loco was limited to 25 kph and still has a little plate in the cab to that effect, so dynamic balancing for high speeds doesn't come into it.

As both the fly cranks and the wheels would have keyways cut into the axles, my hunch is that at an earlier rebuild the keyways for the wheels were found to be worn (unusual but possible if badly fitted keys had been used previously) and new ones cut. New keyways for the flycranks would have been visible from yhe outside but none are apparent. But that is just a wild guess.

It will be interesting to see if the alignment is remedied in the current rebuild, but sadly they are not updating the Facebook page very often or in great detail so we may not get to see.

But well done for spotting it!

Best wishes

Malcolm