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

[don9f]

Hi Roger, thanks for those videos....fascinating stuff indeed. It will be interesting to see the performance in a scale 8x body!

Cheers Don

[Roger]

Hi Don,
Here's the Cone being tested in the scale 8X body. It actually works better than it did in the development body. I'll have to check and see if the Steam cone was made as 1.05mm and not the 1.0mm size I tested it with. That probably accounts for the lower pressure performance.

YouTube link


20210213_210720 by Georgia Montgomery, on Flickr


The scale body need to have the inside of the cap at the delivery end changed, it's still flat inside. I also need to test it with a ball in the clack valve. that might cause problems as currently made because it's pretty tight for room in there.

Anyway, I think I can call that a success at this point. There are a few more things to test, one being the very high temperature performance and the other being whether those overflow pockets at the Delivery end need to be there.

Out of interest, I didn't end up drilling the back of the Steam Cone, it worked with just the hole and taper being created from the front with the back opening into the side cutout where the Steam enters.

[Oily Rag]

Bravo, how could one not be impressed ? Persistence of toil has rewards.

[Roger]

Thanks Dazza, that's kind of you.

So here's the state of play as far as what was demonstrated in the last video.

I still have to double check the size of the Steam Cone hole, but having changed the model to exclude the tapered rear, this is how it looks. I'm pretty sure the throat is 1.1mm which I already showed as too big when I tested it on the Development Injector.

I'm showing this because the back of the Steam Cone is a long way from what is usually done and what I've tested with all along. I really didn't expect this to work because of the long parallel section of throat, and the sharp corner where it meets the cutout. Ok, I can't say I'm surprised that the sharp corner didn't make much difference, Bob Bramson had already said as much.

However, that long throat might have been problematical, but it might just turn out to be a discovery. You saw how this went down to 20psi whereas the Development Injector didn't quite. So the question is, how does a long throat affect the flow rate of the Steam? Does this arrangement reduce the flow more at high pressures than it would if it was a short throat? If it does, this could mean that you can use a larger throat size which passes less Steam at the high pressure end, and then compensates by letting more Steam flow at the lower pressure.

Of course, this may be utter nonsense, but it does make me wonder.

Scale 8X injector sectioned by Georgia Montgomery, on Flickr

[jon38r80]

If I remember my Fluid mechanics correctly you have probably allowed the steam flow to settle down to laminar flow getting rid of turbulence after going through all the bends and fittings it went through to get to that point so that you get a more even flow into your expanding steam cone. It is a long time ago, more than 40 years since I did any of that sort of stuff so I am probably wrong.It's a very long time since I applied Bernoulli to anything and to think I used to be able to derive the equation from first principles.
Im probably talking throuigh my hat anyway.

[simplyloco]

Feb 14, 2021 12:04:58 GMT jon38r80 said:

If I remember my Fluid mechanics correctly you have probably allowed the steam flow to settle down to laminar flow getting rid of turbulence after going through all the bends and fittings it went through to get to that point so that you get a more even flow into your expanding steam cone. It is a long time ago, more than 40 years since I did any of that sort of stuff so I am probably wrong.It's a very long time since I applied Bernoulli to anything and to think I used to be able to derive the equation from first principles.
Im probably talking throuigh my hat anyway.

I could do that too, but the hot air disintegrated my hat some years ago...

[flyingfox]

Yes, it needs 10 times the diameter a fluid is operating in to achieve laminar flow, the more the better, and I think that Roger has demonstrated than.
Keep going Roger, this is fascinating work
regards
Brian

[Roger]

A quick update on finalising this design.

Firstly, here's a better view of the Steam Cone and the extended throat. I've just checked this, and it is indeed 1.1mm which was too big when I tried that size on the development cone. Bear in mind that the development Steam Cone has a long tapered lead in at 13 degrees to a short 1mm diameter parallel throat of say 1mm length.

The fact that this works even better than the 1mm cone on the development Injector tells me that something significant is happening here. Clearly there's more restriction, and that is presumably what's making the difference. I imagine the pressure at the expanding section is slightly reduced, and that's why it works.

Adjustable steam cone section by Georgia Montgomery, on Flickr

I've also tried the Injector today with the 4mm Silicon Nitride ball in the clack, removing the one on the boiler so I know it's the only one doing anything. The result is that it works just fine, even though it does look like it might restrict the flow. It also doesn't appear to get stuck in the cap. I have put a few notches in the conical edge inside the cap to make sure there's no way it would form a seal, so maybe that helped.

And finally, I've made Cone 54 which is identical to the last one with the exception of not having the pockets for the four 0.5mm delivery overflow holes.

So that almost wraps up the development, I don't think there's much more performance to be wrung out of the design. I've got some 1.5mm and 1.2mm diameter Silicon Nitride balls on order to perform some measurements of the tapers. I want to more accurately cross check their positions in the body to the 3D model. Once I've done that, I'll be in a position to document this with confidence and then I'll probably write it up for the Model Engineer magazine.

[Gary L]

Feb 14, 2021 12:48:07 GMT Roger said:

A quick update on finalising this design.

Firstly, here's a better view of the Steam Cone and the extended throat. I've just checked this, and it is indeed 1.1mm which was too big when I tried that size on the development cone. Bear in mind that the development Steam Cone has a long tapered lead in at 13 degrees to a short 1mm diameter parallel throat of say 1mm length.

The fact that this works even better than the 1mm cone on the development Injector tells me that something significant is happening here. Clearly there's more restriction, and that is presumably what's making the difference. I imagine the pressure at the expanding section is slightly reduced, and that's why it works.

Adjustable steam cone section by Georgia Montgomery, on Flickr

[Snip]

The clue might be in Brian’s post; if it works better, it might be because you have a longer parallel section and thus get closer to laminar flow -?

Also isn’t the pressure increased not decreased in the expanding section? Wider section results in slower speed; slower speed means higher pressure & vice versa?

Gary

[Roger]

Feb 14, 2021 13:59:48 GMT Gary L said:

Feb 14, 2021 12:48:07 GMT Roger said:

A quick update on finalising this design.

Firstly, here's a better view of the Steam Cone and the extended throat. I've just checked this, and it is indeed 1.1mm which was too big when I tried that size on the development cone. Bear in mind that the development Steam Cone has a long tapered lead in at 13 degrees to a short 1mm diameter parallel throat of say 1mm length.

The fact that this works even better than the 1mm cone on the development Injector tells me that something significant is happening here. Clearly there's more restriction, and that is presumably what's making the difference. I imagine the pressure at the expanding section is slightly reduced, and that's why it works.

Adjustable steam cone section by Georgia Montgomery, on Flickr

[Snip]

The clue might be in Brian’s post; if it works better, it might be because you have a longer parallel section and thus get closer to laminar flow -?

Also isn’t the pressure increased not decreased in the expanding section? Wider section results in slower speed; slower speed means higher pressure & vice versa?

Gary

Hi Gary,
Approaching laminar flow might help with stability, but if the 1.1mm throat was too big on the development cone, it ought to be too big here.

I see where you're coming from with the cross section being larger, hence a higher pressure, but that would only be true if the original Steam Cone had a similar length of parallel portion. On the development Injector, the parallel 1mm throat section is tiny, so the pressure immediately before the throat is high. On this cone, I suspect the pressure at the start of the long 1.1mm daimeter parallel section is similar to that right at the 1mm throat diameter of the development Injector.

One way or another, it seems to me that the amount of Steam exiting this 1.1mm nozzle is similar to the 1mm nozzle on the development injector. If it was any more, I don't think it would work. If laminar flow was the only thing affecting it, I don't think that would decrease the volume of Steam being passed.

Who knows, it's a whole new avenue of experimentation I don't feel the need to explore at the present time, interesting though it is.

[chris vine]

Hi Gary,

I think that once the nozzle is choked, IE the gas/steam is at the speed of sound, then the diverging part of the nozzle accelerates the flow and the pressure decreases.
Look at the shapes of nozzles on rocket engines.

Chris.

[Oily Rag]

"I'll be in a position to document this with confidence and then I'll probably write it up for the Model Engineer magazine."

YAY. I am keen to one day have a go at making one of these little fellas.
I hope to be able to import the 3D drawings into Solidworks. I have never done that, but if good 2D drawings are available, no problemo

[Roger]

Feb 14, 2021 20:19:22 GMT Oily Rag said:

"I'll be in a position to document this with confidence and then I'll probably write it up for the Model Engineer magazine."

YAY. I am keen to one day have a go at making one of these little fellas.
I hope to be able to import the 3D drawings into Solidworks. I have never done that, but if good 2D drawings are available, no problemo

Hi Daz,
With your eye for accuracy it will be a doddle. You're more than welcome to any of the 3D models when I've verified them against the cones.
In my opinion these are far easier to make than conventional cones, although you do need a way to accurately cross drill holes. Still, even that can be done if you design a simple indexing fixture for a Lathe.
The most difficult part is drilling the deep centre hole, but that's not too bad if you go about it the right way.
Designing a body for them is easy enough.

[jma1009]

Hi Roger,

Does it work feeding hot water?

This was one of your aims.

If it does, does it feed water above 46 degrees centigrade?

Cheers,
Julian

[jma1009]

( By the above I obviously mean the temperature of the water coming into the injector )

[delaplume]

Feb 14, 2021 16:14:07 GMT chris vine said:

Hi Gary,

I think that once the nozzle is choked, IE the gas/steam is at the speed of sound, then the diverging part of the nozzle accelerates the flow and the pressure decreases.
Look at the shapes of nozzles on rocket engines.

Chris.

Surely the basic venturi principle would still apply,  no matter what ??......so diverging ie expanding will slow the velocity and increase the pressure ??....Have a look at the final variable ducting shapes used on Concorde's air inlet side to the Rolls-Royce engines...

[jma1009]

[/quote]Surely the basic venturi principle would still apply,  no matter what ??......so diverging ie expanding will slow the velocity and increase the pressure ??....Have a look at the final variable ducting shapes used on Concorde's air inlet side to the Rolls-Royce engines...[/quote]

When an injector is working properly, it is working at below atmospheric pressure until you get to the delivery cone.

This is why there is traditionally a check valve (or flap valve) at or on the first half of the combining cone, or the sliding first half of the combining cone that Ed intends to have a go at.

Transferring the check valve in effect to the start of the delivery outlet, as Roger had done, could have been done by turning upside down a standard injector and removing the ball from the combining cone check valve, and adding (upside down) a check valve to the outlet to for the overflow from the gap between the end of the combining cone and start of the delivery cone.

I would question that Roger's model is easier to make. I am quite happy with press fits and delivery cone and combining cone 2 halves as 3 separate pieces, and I remain unconvinced that 'end regulation' is easier than the annular gap for the steam cone just into the start of the first half of the combining cone.

[coniston]

That's superb work Roger and a wonderful achievement. It certainly looks to be a well working injector with a good range of operating pressure. One thing from you videos you don't show starting the injector at lower pressures from 'cold' only restarting when pressure falls to around 40PSI. I assume it will start dry at any pressure in between? I don't see any reason for it not doing so, just not seen it demonstrated in your videos.

Excellent

Chris D

[Roger]

Feb 14, 2021 20:51:32 GMT jma1009 said:

Hi Roger,

Does it work feeding hot water?

This was one of your aims.

If it does, does it feed water above 46 degrees centigrade?

Cheers,
Julian

Hi Julian,
I'll run some more tests with hot water shortly. So far the designs I've tested work with feed water up to about 35C, but I'll have to explore the limits with this one. I don't suppose it will be any different, but I need to check.
Unlike most Tank Engines, mine will feed water exclusively from the Tanks when using the Injectors because that's the way they're plumbed on 1501. That being the case, there will usually be a constant flow from the tanks, even if water is being drawn from the riding truck. I very much doubt if these will ever see 30C in service, but you never know!

[chris vine]

Hi Alan,

Once the speed of the gas reaches Mach 1 (speed of sound in the gas at its particular conditions) at the throat of a convergent/divergent nozzle, it is said to be choked.
After that, it accelerates in the divergent section of the nozzle, just like in a rocket motor, and the pressure falls.
If the flow at the throat is less than Mack 1, then as you say, the flow will slow down in the divergent section and the pressure will rise.

In the injector you are trying to get the maximum velocity and hence momentum into the steam before it hits the wall of cold water and condenses. It is the convergent / Divergent nozzle which achieves this.

Much more about them here from Nasa www.grc.nasa.gov/www/k-12/airplane/nozzled.html but the critical bit is copied here: For our CD nozzle, if the flow in the throat is subsonic, the flow downstream of the throat will decelerate and stay subsonic. So if the converging section is too large and does not choke the flow in the throat, the exit velocity is very slow and doesn't produce much thrust. On the other hand, if the converging section is small enough so that the flow chokes in the throat, then a slight increase in area causes the flow to go supersonic. For a supersonic flow (M > 1) the term multiplying velocity change is negative (1 - M^2 < 0). Then an increase in the area (dA > 0) produces an increase in the velocity (dV > 0). This effect is exactly the opposite of what happens subsonically.

Chris.
PS, Alan, you mentioned the intakes on Concorde. They were very special to deal with both subsonic flow at take-off and then supersonic flow once flying supersonically. I was told years ago by an old member of staff at Rolls Royce Bristol that when Rolls Royce bought the Bristol (siddley?) engine company, the only thing the RR guys were really interested in was the Concorde intakes.