Experimental Flash Steam and others

[Richard Hed]

Ajoeiam, re post #245: Distortion after welding is usually a combination of the size/stiffness of the parent part, and the amount of weld and heat input.
Simply: The weld is molten, so when it "freezes" and becomes attached to the already "frozen" parent metal, it is cooling from a state that is hotter than the parent metal. The differential contraction as parent metal and weld metal cool cause internal stresses at the weld interface... and if great enough to distort the parent metal it is these stresses that cause distortion. Uneven cooling can make that condition worse, as we tend to cool parts with quenching "the far end" of the part, but this can cool the parent metal faster than the weld and exacerbate a problem that is inherent in the welding process. Pre-heating and normalising are simply ways to reduce weld stresses, thus reducing distortion. But microscopically, all weld repairs distort parent metal because we lay and bond molten metal onto cooler "frozen" metal. It is just that we can get away with some very small distortion when it is so small it is within tolerance.
K2

That's a great idea to use the lathe as a mill to cut the keyway.

I was thimpfking there might be several ways to reduce the heat distortion. One wojuld be to pein it, usually done on cast iron, however, it may actually work on steel but I don't know for sure. Another way would be to make very short welds, allow to partially cool before making another. and continue like that till the weld area is very shallow. It is already shallow, at the deep end it is only about 1/8th" deep but less than that wide. and it all narrows down in about an inchand a half. Truthfully, I am not familiar with these types of things. When I weld it is for structures.

I'm thimpfking that a combo of these things might be the way to go, and also that the weld area itself might be small enough to not be too serious. Any ideas?

 

[Steamchick]

Hi Richard,
Yes, I was expecting the grooves to be filled as shallow and narrow. So I would use my brazing hearth and blow-torch and bronze filler rod to fill the grooves. Then allow to cool slowly... I assume the steel shaft for repair is not heat treated, just a "good steel" as it will have been a ground finish as new.
Distortion should be minimal, as steel contraction of the shaft versus the thin section of bronze will not have a lot of residual stresses (different coefficients of expansion). I think filling with bronze is practical as the damage is shallow, so not affecting the machine's strength significantly. Even silver solder should work, just not so hard in service if the shaft has things rubbing on it - like plain bearings? Which is why I suggest a Bronze brazing rod filler.
Alternatively, grind out the groove to a uniform "Vee""-ish shape so a single uniform weld the full length will fill the groove in one pass. Then the only slag will be on the surface that will be machined or filed off. (if using stick welding. MIG has no slag). The amount of heat introduced will not be great. And relative stresses will be a weld of 1/8" approx triangle versus the original steel shaft steel cross section (thick walled tube?). so only a few % difference, so hardly any distortion?
I wish I was able to be there to help.
K2

 

[ajoeiam]

That's a great idea to use the lathe as a mill to cut the keyway.

I was thimpfking there might be several ways to reduce the heat distortion. One wojuld be to pein it, usually done on cast iron, however, it may actually work on steel but I don't know for sure. Another way would be to make very short welds, allow to partially cool before making another. and continue like that till the weld area is very shallow. It is already shallow, at the deep end it is only about 1/8th" deep but less than that wide. and it all narrows down in about an inchand a half. Truthfully, I am not familiar with these types of things. When I weld it is for structures.

I'm thimpfking that a combo of these things might be the way to go, and also that the weld area itself might be small enough to not be too serious. Any ideas?

Have not run into any specifications for peening much of anything except on cast iron.
There are vibratory gizmos available for reducing the stress in weldments - - - dunno if any are rated for something this small - - - sorry.
Using a decent welding procedure (welding round and round and working to keep slag inclusions LOW and the heat up (fusion to both the parent shaft and the other build up material) is usually enough to minimize any problems.

Now if you insist on 1" long welds and spend a lot of time chipping in between - - - imo - - - you are far more likely to actually create issues but its your show - - - - the doing and the joys afterward are all 'yours'.

 

[ajoeiam]

Hi Richard,

snip

Which is why I suggest a Bronze brazing rod filler.
Alternatively, grind out the groove to a uniform "Vee""-ish shape so a single uniform weld the full length will fill the groove in one pass. Then the only slag will be on the surface that will be machined or filed off. (if using stick welding. MIG has no slag). The amount of heat introduced will not be great. And relative stresses will be a weld of 1/8" approx triangle versus the original steel shaft steel cross section (thick walled tube?). so only a few % difference, so hardly any distortion?
I wish I was able to be there to help.
K2

If all you're comfortable with is brazing - - - - braze away.
MIG has slag - - - - just not very much of it.
Dunno about you but if I were using 3/32" welding rods I think I would have something like a 3/16" and maybe even a little closer to a 1/4" fillet size - - - that is not large.

Given the conversation - - - - it might be easier to just buy some 1045 T&G (turned and ground) to the size needed and then to machine in the keyway using the poor shaft and then do a switcheroo (you could also machine the keyway in the lathe - - - up to you!).
Then you have no worries about HAZ or distortion or or or.

 

[Richard Hed]

Have not run into any specifications for peening much of anything except on cast iron.
There are vibratory gizmos available for reducing the stress in weldments - - - dunno if any are rated for something this small - - - sorry.
Using a decent welding procedure (welding round and round and working to keep slag inclusions LOW and the heat up (fusion to both the parent shaft and the other build up material) is usually enough to minimize any problems.

Now if you insist on 1" long welds and spend a lot of time chipping in between - - - imo - - - you are far more likely to actually create issues but its your show - - - - the doing and the joys afterward are all 'yours'.

On no account will I use stick welding. MIG or TIG. I keep wondering what you mean by round and round--is this the welding technique?

 

[ajoeiam]

On no account will I use stick welding. MIG or TIG. I keep wondering what you mean by round and round--is this the welding technique?

Why not stick - - - - crap - - - its as easy peasy as eating pie!!!

Just mean that you are turning the shaft as you're welding it so the weld is a continuous spiral.
You only stop welding when your rod is too short or you are getting too hot or you can no longer turn the shaft.
(If rod is too short - - - hit the weld to knock off the flux drop in a new rod and git welding!)

 

[Richard Hed]

Why not stick - - - - crap - - - its as easy peasy as eating pie!!!

Just mean that you are turning the shaft as you're welding it so the weld is a continuous spiral.
You only stop welding when your rod is too short or you are getting too hot or you can no longer turn the shaft.
(If rod is too short - - - hit the weld to knock off the flux drop in a new rod and git welding!)

That gouge is not as large as it looks in the photos--so going round and round would be putting a lot of metal on places it is definitely not needed. It only needs some added metal in a slim spot on the edge of the keyway. I am not willing to risk all that slag in the corners that I cannot get out and especially when it rattles loose in the machine when it is put back into service.

Yes, I agree with you about stick welding is easy peasy, but not for this particular job. I am leaning towards

955 Aluminum Bronze​

brazing which has all the needed characteristics of toughness, hardness, strength, and lubrication. It also is a lower temperature to braze than to weld which I consider an important topic.

To wreck that engine keyway had to take a hell of a lot of energy. I suspect it didn't happen all at once and that the operator did not know or notice the problem. I do not know if it would make a noise in doing so, but I would suspect it would have noise of some kind or at least a different sound especially on startup and stopping. I thimpfk that probably any solution would work, as there is not really that much at work here, it just needs to be accurately done and the slot finished correctly. However, there is for certain, a "best" solution. I'm leaning toward that 955 alum bronze. Any suggestions?

I really thimpfk that the design of the key is terrible. It is a 1/4" plastic holder with a 3/32 or so steel piece embedded in it. All this has a screw hole in the middle to keep it together and centered in the sliding keyway. this makes the walls at the screw hole extremely weak--bad design--and this is where it broke of course. On utub there are lots of examples of this breaking at that point. There must be a better solution.

 

[ajoeiam]

That gouge is not as large as it looks in the photos--so going round and round would be putting a lot of metal on places it is definitely not needed. It only needs some added metal in a slim spot on the edge of the keyway. I am not willing to risk all that slag in the corners that I cannot get out and especially when it rattles loose in the machine when it is put back into service.

Yes, I agree with you about stick welding is easy peasy, but not for this particular job. I am leaning towards

955 Aluminum Bronze​

brazing which has all the needed characteristics of toughness, hardness, strength, and lubrication. It also is a lower temperature to braze than to weld which I consider an important topic.

To wreck that engine keyway had to take a hell of a lot of energy. I suspect it didn't happen all at once and that the operator did not know or notice the problem. I do not know if it would make a noise in doing so, but I would suspect it would have noise of some kind or at least a different sound especially on startup and stopping. I thimpfk that probably any solution would work, as there is not really that much at work here, it just needs to be accurately done and the slot finished correctly. However, there is for certain, a "best" solution. I'm leaning toward that 955 alum bronze. Any suggestions?

If all you do is fill the depressions using either brazing or tig or any other process - - - - I think you are most likely to introduce stresses into that shaft which are most likely to show up as distortion.

If you are only intent on something like sticking a little something into the wrinkles - - - - I would use something like an epoxy or locktite had a green bushing
retaining compound. Glop that on, install putting in a decent key and it will likely outlast your work needs.

(I worked for an outfit where they even did ceramic infills - - - occasionally (IIRC) on rotors that were used at around the 8k rpm range. You might be surprised at how repairs are done even on some very high powered high load very high duty cycle equipment.)

All the best in your fix.

 

[Richard Hed]

If all you do is fill the depressions using either brazing or tig or any other process - - - - I think you are most likely to introduce stresses into that shaft which are most likely to show up as distortion.

If you are only intent on something like sticking a little something into the wrinkles - - - - I would use something like an epoxy or locktite had a green bushing
retaining compound. Glop that on, install putting in a decent key and it will likely outlast your work needs.

(I worked for an outfit where they even did ceramic infills - - - occasionally (IIRC) on rotors that were used at around the 8k rpm range. You might be surprised at how repairs are done even on some very high powered high load very high duty cycle equipment.)

All the best in your fix.

Yes, epoxy I considered briefly and I will reconsider it more now that you have told me about this stufff that is done in industry. I thimpfk that almost any repair would work if fFI could solve that ****** engineering problem of the hole in the sliding key. I saw one solution for a different placement in Bridgeports--not the motor section, but the pulley rotor at the other end of gthe head, in which a plastic (delrin or nylon or something strong but with good lube properties) in which a piece was machined to fit in the slot, however, a round *** was placed where the screw hole goes, all one piece. this was then epoxied in for solid fit. I'm considering this, however, since it is the power side, I am not so sure.

As you say, tho', it will proabably out last me. Even if it didn't keep for a long time, it might keep long enough, say a couple years, to simply replace it now an then. taking the head apart is quite easy. I compare that to removing an engine from a car, MUCH easier. With plastic, it is not likely to rip holes in the metal like the last one did. This part only has sideways thrust not up and down the keyway. the only time there is any change in the keyway placement is when one is changing the speeds, and that will not create the kinds of forces that actual milling will create.

Just as an aside, I would thmpfk Bridgeport would be ashamed to make a part that breaks so easily. It's almost as if they got a case of "planned obsolescence".

 

[Richard Hed]

Yes, epoxy I considered briefly and I will reconsider it more now that you have told me about this stufff that is done in industry. I thimpfk that almost any repair would work if fFI could solve that ****** engineering problem of the hole in the sliding key. I saw one solution for a different placement in Bridgeports--not the motor section, but the pulley rotor at the other end of gthe head, in which a plastic (delrin or nylon or something strong but with good lube properties) in which a piece was machined to fit in the slot, however, a round *** was placed where the screw hole goes, all one piece. this was then epoxied in for solid fit. I'm considering this, however, since it is the power side, I am not so sure.

As you say, tho', it will proabably out last me. Even if it didn't keep for a long time, it might keep long enough, say a couple years, to simply replace it now an then. taking the head apart is quite easy. I compare that to removing an engine from a car, MUCH easier. With plastic, it is not likely to rip holes in the metal like the last one did. This part only has sideways thrust not up and down the keyway. the only time there is any change in the keyway placement is when one is changing the speeds, and that will not create the kinds of forces that actual milling will create.

Just as an aside, I would thmpfk Bridgeport would be ashamed to make a part that breaks so easily. It's almost as if they got a case of "planned obsolescence".

I'm curious what other people thimpfk about epoxy. You probably can't see it well enough in the photos, but the gouge is slanted upwhard approx. 45deg. That would make breaking forces that epoxy might not be able to hold. What do you thmpfk?

Maybe with a small amount of undercutting, say with a dremel, it would hold better?

 

[Zeb]

Epoxy becomes brittle after reaching a certain thickness. Even the nasty Hysol stuff.
If you zap it with a grinder, does it shoot a lot of sparks? Welding should do the trick. Brazing might if it don't get crashed again. It might have been good that the plastic part broke because Murphy will always want to look for the next opportune place.

 

[Richard Hed]

Epoxy becomes brittle after reaching a certain thickness. Even the nasty Hysol stuff.
If you zap it with a grinder, does it shoot a lot of sparks? Welding should do the trick. Brazing might if it don't get crashed again. It might have been good that the plastic part broke because Murphy will always want to look for the next opportune place.

Thanx for that. This is not very thick, only about an 8th at the thickest.

 

[Richard Hed]

I managed to buy some silicon brass today to test out for brazing. I am not going to use SiB as Aluminum Brass is much stronger. I have some ordered, supposed to come this Wednesday (a week). I wanted to get a bit of a head start on practice for the AlBrass, even tho' the welding techniques are different. It seems to be OK.

 

[Zeb]

Back on topic, ehem, I decided to share my favorite calculators because I didn't want to trigger a flash steam event on the other thread. I've never used a sliderule....yet.

My all time favorite below. Mine didn't have the nifty carbon fibre though. Didn't really need more than this for setting up a sine plate or laying out bend setbacks.

I ythinkpf what's below doesn't count as calculators, but I call them so... Some workflows could be applied to engine making.

For visualizing more complex algebraic equations, I've used Processing (free) and other interfaces in the past. It does a great job and the custom calculators you can make can be interactive. The two below were done with DSP Robotics using visual programming but processing will do it as well. I like being able to interact with math to understand it better. I'm not math smart enough for an engineering degree, I just made these a while ago when working through a book.

I've tried a bit of matlab, but I never really needed (or understood) the level of math it could produce. Then there's always expressions available in CAD or other 3D programs.

I also tried getting a sliderule on ebay years back but got scammed. The dream died, but maybe I'll pick one up, work through it, and add it to what my wife calls my nerd collection.

 

[Steamchick]

Hi Zeb, I appears you are cleverer than I, so don't discount your abilities.
1: On Slide rules: They are simple scales for adding and subtracting logarithms with scales in "real" numbers.
They do not add, subtract regular numbers. Because of that you have to manually write-down what you are doing, and split calculations into their sub-components, then utilise the slide rule to multiply, etc. by adding the logs of those numbers, etc. That keeps the Slide Rulers directly in touch with their calculations in a way that no computer really does. It is a great tool to help you quickly do the calculations, without taking the calculation into a black-box and just giving you an answer.
2: When you understand what you are doing, you then can write it onto and Excel spreadsheet, programable calculator, or whatever and repeat the calculation many times, even using finely changing variables to iterate to an optimum design.
This is where slide rules become obsolete, although a much speedier way of helping you when doing the basic calculations a smallish number of times.
3: Then (the really clever bit!) you can turn the program inside-out to write it so it will iterate to the optimum, so you input the answer, and it tells you the parameters to get there!

I am at the "Stage 2 level" above, but have had "Mathematicians and computer boffins" write software for stage 3, in my various jobs of a few decades ago.
That was what we called Computer Aided Design before PCs were invented... (Apple, Commodore and the rest!). The programme was on a dedicated huge magnetic disc cartridge, and ran on a computer that was like 10 metal wardrobes in a large room. (Like a school classroom). It took 20 minutes to set-up, then we were allowed 3 runs of 15 mins max. (in the middle of the night). to do the calculations. But it cut a 4 year design and development into 2 years, so was easily justified.
Now you could use your "phone" and get an answer in seconds, if you had "the app"! But you need not understand how the answer was produced... Is that progress?
K2

 

[Zeb]

I will be trying to pick up a slide rule again.
I have a friend that used to write assembly language. I'm sure making a calculator would not be fun, but at 64 bits, I can count on my gorilla resolution to get me past worrying about rounding and floating point numbers. At least we can hope so when calculating interference fit to the 12th decimal.

Here's a "custom" calculator I've worked on today to learn Blender geometry nodes (free) and to wrap my head around Don's very good magneto paper. All the displayed numbers and geometry will update based on the size of magnet typed in. I was playing around with mapping as well, so the RPM selected modifies the energy output based on a (made up) curve. The curve in the bottom right modifies the output based on phase angle. Blender was able to handle all 17,000+ turns of generated coil (very dense black) without any lag.

 

[Richard Hed]

Hey Steam Chik,
I was just perusing "The Una-Flow Steam engine" by Johannes Stumpf. It seems to be something you would like. It is unneccessarily technical for me (but still a good reference book), but seems like it would be right up your line. first published 1912. I have a new copy from "Scholar Select" series. SXCV

 

[Zeb]

Here's a link to it with very nice scans on archive.org. The una-flow steam-engine : Stumpf, Johannes, 1863-1936 : Free Download, Borrow, and Streaming : Internet Archive

 

[Richard Hed]

Here's a link to it with very nice scans on archive.org. The una-flow steam-engine : Stumpf, Johannes, 1863-1936 : Free Download, Borrow, and Streaming : Internet Archive

Ah, that's very nice. My copy is nice except for the print is very tiny. I get a lot of books from Internet Archive, but if possible (read that as "cheap") I prefer paper so's I can make notes in it.

 

[Steamchick]

Thanks Richard, Zeb.
I'll add it to my collection of "reading jobs"!!
But mostly these texts are used as reference books, rather than bedtime stuff.
I was writing to a nephew today - about how I used an 1837 book as a lad to learn the game of Chess. Still have it - it was inherited from my Grandfather - who loved to buy old second-hand books when he was a lad in the late 19th century... So the 1837 text was old-ish (cheap) by his standards. But still just as relevant today. Just like this second edition 1922 Stumpf!
Love it!
K2