Help with 'Speedy' valve gear and other issues...

[kipford]

Roger
If I was doing this professionally on jet pumps I would have a very small flat at the nozzle throat and use a reamer to set that. Keep the flat around 0.25/0.5 mm. It will have no effect on the performance of the injector. It is the throat diameter that is critical here, remember any change in diameter is a square law effect on velocity and pressure in the injector. Small changes in diameter can have a huge effect on performance. Keep up the good work though.
Dave

[Roger]

Feb 3, 2021 22:46:12 GMT kipford said:

Roger
If I was doing this professionally on jet pumps I would have a very small flat at the nozzle throat and use a reamer to set that. Keep the flat around 0.25/0.5 mm. It will have no effect on the performance of the injector. It is the throat diameter that is critical here, remember any change in diameter is a square law effect on velocity and pressure in the injector. Small changes in diameter can have a huge effect on performance. Keep up the good work though.
Dave

Thanks for that Dave, that was my gut feeling. That's what I've done for the Steam Cone, and I think I'll change the design on the One Piece Cone too. Still, it was worth reworking the reamers, I don't like it when something doesn't make sense. Hopefully the errors will be smaller now.

[steamer5]

Hi Julian,
Thanks for the informative reply. I will have to go to my Dads & is if I can borrow my book back and get my head around the finer details!!

Dad is keen on the idea of building an injector, I’ll have to give him a hurry up as he’s 94, hence the reason for borrowing my book.

Cheers Kerrin

[Roger]

This is the adjustable Steam Cone adaptor for the scale 8X Injector body. I'm making this prototype in Brass for ease, the ones for the Locomotive will be Phosphor Bronze and use a finer thread.

20210203_125248 by Georgia Montgomery, on Flickr

I screwed this tightly into the body and then inched the 4th axis round until it looked vertical.

20210203_153639 by Georgia Montgomery, on Flickr

The Steam and...

20210203_155527 by Georgia Montgomery, on Flickr

... the Water inlet pockets were then cut on opposite sides. I'm really liking this ER32 collet adaptor in the 4-Jaw. It's really quick to set up and it gives more clearance for the spindle.

20210203_161320 by Georgia Montgomery, on Flickr

That was screwed tightly into my standard M7 x 0.75 (fine) sleeve and the inside details machined.

20210203_163845 by Georgia Montgomery, on Flickr

The adjustable Steam Cone has to be Phosphor Bronze because it would be too weak in Brass. Once again the 4th axis came in really handy, with just one program being required for the pocket and the comment ANGLEREP4 added to the output so I could create the indexed program with just the press of a button. This really has been well worth the effort to develop, I'm using it all the time.

20210203_201644 by Georgia Montgomery, on Flickr

I screwed the parts tightly together and then gingerly machined both hex heads 0.1mm at a time to make sure they didn't unscrew.

20210203_215240 by Georgia Montgomery, on Flickr

They're pretty small and a bit fiddly to make, but the benefits of being able to adjust this are enormous.

20210204_120905 by Georgia Montgomery, on Flickr

For the moment, I haven't drilled out the back of the cone and added the tapered lead in. The throat size goes through to the big cutout, so there's still a connection. I thought I'd try it like this to see if it will work anyway before doing that.

20210204_120844 by Georgia Montgomery, on Flickr



20210204_120549 by Georgia Montgomery, on Flickr

I don't think it's obvious that the small screw isn't a flanged cover like the real thing. It's a small compromise for something that useful.

20210204_123043 by Georgia Montgomery, on Flickr

[delaplume]

Hi Roger,

Could you post a photo showing this injector in the palm of your hand please ??.........Just by way of a reality check to remind us again just how small this is ...Many thanks ---------Alan

[Roger]

The Tool and Cutter Grinder has this little handwheel for moving the swivel table, which can be quite stiff. You move is so far, then release a driving dog, return the rack to a new position and re-engage the dog. It's a good system for fine adjustment, and you can move the table by hand a long way if you force it. However, I've never been happy with it, the knob needs far too many turns when you need to keep changing the angle, such as when grinding reamers.

20210204_151113 by Georgia Montgomery, on Flickr

Fortunately, there's 25.4mm parallel section to the inside of the knob, so this is an 8mm hex being machined into a piece of Steel bar to fit inside it.

20210204_152352 by Georgia Montgomery, on Flickr

It was a nice close fit, so I just secured it with Loctite. Now I can use the crank handle that has an 8mm hex ball on the end, the one I made for rapidly adjusting the vice on the Power Hacksaw. That's made life so much easier, it's a pleasure to use instead of a chore.

20210204_174028 by Georgia Montgomery, on Flickr

[jma1009]

I am a bit bemused how Roger has got some 'throat' sizes departing from his 3rd January throat sizes of steam cone 0.040", combining cone 0.032", and delivery cone 0.026".

The delivery cone throat is the most crucial, and in Roger's one piece combining and delivery cone assembly must obviously occur at the correct location.

When setting everything up for drilling and reaming I drill a size or 2 under and ream the second stage till the end of the next drill size below final size is sort of a push fit on the drill shank on the cone; I then drill through with a proven drill of the final throat size, that also provides a short parallel section mentioned by 'Kipford'.

If inadvertently I've made a mistake with the final reaming and gone a bit too far that bit of brass goes in the scrap bin. I check everything as I go along but you do get errors sometimes with a bit of swarf find it's way onto the end of in my case the lever tailstock, or simply applying too much pressure to same or just getting a bit tired!

I don't have Roger's super duper readouts, or even the stops for a capstan tailstock on Arthur Grimmett's old Atlas lathe; so for me each bit gets inspected and checked individually to pass Arthur's "window test".

Anyway, if Roger has made his single one piece combining cones and delivery cones and not got the delivery cone throat the right size and not within the design parameters then I'm not surprised that his results have been rather random and not as expected so far!

Quality control?!

Cheers,

Julian

[JonL]

Do you ever read your posts before you send them? I know you are attempting to be constructive but they often come across as downright rude. Quality control?!

[Roger]

Feb 4, 2021 20:35:08 GMT jma1009 said:

I am a bit bemused how Roger has got some 'throat' sizes departing from his 3rd January throat sizes of steam cone 0.040", combining cone 0.032", and delivery cone 0.026".

The delivery cone throat is the most crucial, and in Roger's one piece combining and delivery cone assembly must obviously occur at the correct location.

When setting everything up for drilling and reaming I drill a size or 2 under and ream the second stage till the end of the next drill size below final size is sort of a push fit on the drill shank on the cone; I then drill through with a proven drill of the final throat size, that also provides a short parallel section mentioned by 'Kipford'.

If inadvertently I've made a mistake with the final reaming and gone a bit too far that bit of brass goes in the scrap bin. I check everything as I go along but you do get errors sometimes with a bit of swarf find it's way onto the end of in my case the lever tailstock, or simply applying too much pressure to same or just getting a bit tired!

I don't have Roger's super duper readouts, or even the stops for a capstan tailstock on Arthur Grimmett's old Atlas lathe; so for me each bit gets inspected and checked individually to pass Arthur's "window test".

Anyway, if Roger has made his single one piece combining cones and delivery cones and not got the delivery cone throat the right size and not within the design parameters then I'm not surprised that his results have been rather random and not as expected so far!

Quality control?!

Cheers,

Julian

Hi Julian,
It was a pretty minor error on the model that resulted in the second taper not opening out the throat far enough. I'd naively assumed that the care I'd taken with everything would result in achieving the right size. Lesson learned. I've now got some gauges made and I'll change the design and not just go straight to the sizes without checking the throat. The reamers appear to produce slightly over sized tapers, and that's probably due to setting the depth on a sharp edged hole. I'll have to look at other ways to do that more accurately.

It's easy to be wise with hindsight, especially when your focus is on other elements of the process.

[Oily Rag]

R , snipped "It's easy to be wise with hindsight, especially when your focus is on other elements of the process."

Hear Hear.

[Roger]

I've carefully opened out a few of the more promising cones that had been made with the throat too small, and that led me to think that this style of cone is worth a little more investigation.

You can't really corrolate this with the usual design nomenclature of Condensing Cone and Mixing cone, because their functions are blurred. You can't relate any of the usual throat dimensions, other than the Delivery Cone throat.

If you look at the full size 8X or 10X injectors they open the whole side of the snout, and when that closes, there's no way it's going to seal along the sides. We've also seen that Steam lance design that has a similar pattern to what I've shown below.

Anyway, this has 4 x 0.5mm holes in rows 1,2,3 & 5 in an attempt to keep them from interfering with each other.

I've just given this a try, and with 20C water it was dry from 90-30psi when it dropped out. I could restart it and get it down a little more by restricting the water but not right down to 20psi.

This is quite promising, it's the best result so far. I think it might benefit from making the first row of holes slightly bigger. Even so, it's starting and restarting well, so another iteration is in order.

I'm still not convinced that trying to simulate a straight edge with overlapping holes gives a better result. Making a single gap the size it needs to be seems to cause problems. This might be the solution. Ideally I'd like to keep the holes all the same size. However, they get closer together the further downstream you get.


Cone27 by Georgia Montgomery, on Flickr

[dscott]

For the many who follow Rogers progress on here and in our various meet ups in Real Life.
As Alan asked for a scale here is the wonderful brake fitting shown to the Faithful at our last 2020 meet up
before it all stopped in Rogers hand. Alexander Palace 2020.



Joking aside we all bring something we have been working on to hand round during.
Our next meet up may include finished models it has been so long!!

David and Lily.

[Roger]

Just a few observations about what I've noticed going on when the Injector is running. We're all familiar with the chirping noise that's associated with an injector doing it's thing, and I think it's generally considered to be a good sign. Bob Bramson describes this as an indication of instability, and I have to say that I agree with him. I think the sound is the ball valve slamming up and down on the seat, as a little excess water escapes to let the flow stabilise.

I've discovered that the dynamics of delivering water to the boiler is very different to that of feeding water to the Balanced Clack valve which is very sensitive to instability in the delivery. This manifests itself as both a noise and also that's usually accompanied by a reduced delivery and the overflow becoming wet. However, just turning the valve to let it feed the boiler, usually results in the flow becoming smoother and the overflow dry.

I can usually tell if it's going to be a more or less successful test based on how it sounds and how it delivers into the Balanced Clack valve without even bothering with feeding the boiler. The quieter and drier the overflow with that valve, the better it will perform when feeding the boiler.

If I'm conducting a dropping pressure test, feeding the boiler with the heater off, I can hear the instability change as the pressure drops. It always seems to be unstable and noisier at the start, becoming smooth and quiet as the pressure drops. In fact it gets so quiet that I'm not aware that it's feeding at all. It seems that the sooner it goes quiet as the pressure drops, the sooner it will drop out. If this sound was just the general noise it makes because it's water flowing at speed, that wouldn't change between cone designs.

Hopefully this will become clearer as I progress, but it's surprising just how differently similar cones perform and sound. Subtle changes make big differences, usually for the worse!

[delaplume]

You're on the right track there, Roger...........On the footplate you can definitely hear the feed-water "singing" when the injector is working-- even with the various other sounds of a loco on the move...

On a model injector I was of the opinion that the "chirping" noise....and it does indeed sound like a small bird chirping --- was air being draw in via the overflow ..... If it was the ball valve slamming down wouldn't that have more of a staccato quality to it ??........In either case, contrary to the "Wisdom of Ages" that has to be detrimental I would have thought ??

Crack-on matey !!........I might not post much but always read what you do and learn from it ( However, I don't think that IBM will be Head-hunting me as a Chief Programmer anytime soon ---LoL !! )...

[Roger]

Feb 5, 2021 10:57:47 GMT delaplume said:

You're on the right track there, Roger...........On the footplate you can definitely hear the feed-water "singing" when the injector is working-- even with the various other sounds of a loco on the move...

On a model injector I was of the opinion that the "chirping" noise....and it does indeed sound like a small bird chirping --- was air being draw in via the overflow ..... If it was the ball valve slamming down wouldn't that have more of a staccato quality to it ??........In either case, contrary to the "Wisdom of Ages" that has to be detrimental I would have thought ??

Crack-on matey !!........I might not post much but always read what you do and learn from it ( However, I don't think that IBM will be Head-hunting me as a Chief Programmer anytime soon ---LoL !! )...

Hi Alan,
I guess it depends on where the air is being drawn in. There's no valve on the area where the Delivery Cone overflow is, and I doubt very much that what we hear is anything to do with that area.
If it's the ball valve, which I think is the case, then it has to move to allow air in, and if the flow is stable, then the ball should be properly seated. Since you hear this effect on every Injector (as far as I know), then you'd have to assume that at least some of those Injectors have properly seating balls the shouldn't let air past them.
So my conclusion is that the ball has to be moving off the seat, and the only thing that can make that happen is something pushing up from below, and that's water.

I reserve the right to be completely wrong, it's just how I imagine it!

[John Baguley]

Same as Alan, I've always been under the impression that the 'chirping' is air being drawn into the overflow caused by the jet from the combining cone not completely filling the diameter of the delivery cone. It then acts like an ejector and sucks air in. Not too bad with a copper boiler but not something you want in a steel boiler! Probably something I read in one of the many articles I studied when I was playing about with injectors.

John

[Roger]

Feb 5, 2021 11:20:33 GMT John Baguley said:

Same as Alan, I've always been under the impression that the 'chirping' is air being drawn into the overflow caused by the jet from the combining cone not completely filling the diameter of the delivery cone. It then acts like an ejector and sucks air in. Not too bad with a copper boiler but not something you want in a steel boiler! Probably something I read in one of the many articles I studied when I was playing about with injectors.

John

Hi John,
I suppose that ought to be easy to prove. if injectors with an additional valve on the overflow don't make that noise.
I guess that drawing air into the boiler at the delivery is unwelcome in any case, because it reduces the thermal efficiency of the injector, the air cooling the water. I also understand that this is another reason why injectors don't like hot water because any entrapped air makes the column of water somewhat elastic.
I'd be interested to know the answer for certain.

Actually, removing the ball ought to prove the point. It will still work after a fashion without the ball

[Oily Rag]

With said injector running/chirping, would not placing one's finger over the over flow pipe end determine if it is air that is going in ?
Yeah, I admit I am dim if I am off track.

[jma1009]

As for being to "rude" yesterday evening to Roger I apologise to Roger.

What other forum members may not appreciate is that Roger and I have met each other on 4 occasions in the last 6 years or so, and for more than 6 years have also had at times considerable email correspondence.

Roger knows that I am blunt, and he had to redo the rear bunker. But it is a 2 way process.

Anyway, apologies, Roger, for any offence caused by me.

Cheers,

Julian

[Roger]

So today I've had to make another batch of Injector bodies because I'd used all the ones I made last time!

While they were being made, I did some experiments to make sure I know the correct setting length with the new reamers. It turns out that they enter the setting piece further than I thought, and that combined with cutting over size results in needing to set them with a 13.5mm Slip Gauge instead of the 14mm I was using. That may be very slightly too much, but better that way than too big. The idea from now on is to create the tapers from both ends and then drill out the small parallel section afterwards to be sure of getting the right size. The drill won't have much to remove, so it ought to cut to size.

20210205_211442 by Georgia Montgomery, on Flickr

Here are another 18 cones ready for the next adventure!

20210205_213402 by Georgia Montgomery, on Flickr